Skip to main content

Full text of "Electrical age [microform]"

See other formats

xf — icx twp X 





^PfeiYf' i 


^i f 

k & b^VJI 


"N m&Bl 


M : 

* , '^f 







!* O- .- ,V 


Mr. <5 1 i •. 


fa $>- 

^ * ^mP : 








30 JUN 1931 




January— December, 1908 





Public Service Commission, New York i 

Power Rating of Generators i 

The Tungsten Filament i 

A New Storage Cell _> 

The Central Station Distributing System 2 

Personality 2 

The Estimation of Power-Factor on Unbalanced Three- 

Phase Loads 3 

The Central Station Distributing System-Feeder Calcula- 
tions 5 

Modern Arc Lighting 9 

Systematic Testing oi Oil in Transformers, and Methods of 

Reclaiming Oil for Ser\ ice 13 

Single-Phase Textile Motor Installation 15 

The Rosenberg Generator 17 

Tungsten Series Incandescent Lamps at Grosse Point. Mich.. 18 

Electric Locomotives — Continued 10 

Electric Motor Connections 2? 

Selling Currents 

Questions and Answers 27 

Review 1 >f the Technical Press 

Westinghouse Reorganization Plan 29 



The Professional Electrical Course 3J 

The Westinghouse Situation 37 

The Tungsten Filament t,- 

The Step Bearing 37 

The Central Station Distributing System 38 

Electric Locomotive Rail Pressure 43 

French Underground Cables 4,5 

A Peculiar Turbine Trouble 44 

Electrolysis 45 

An Odd Case of a Motor Dropping Its Load 53 

The Application of Electric Power to Pulp and Paper Mills. . 54 

The Westinghouse Electrically Heated Sad-iron 60 

The Electric Drive in a Hardware Factory 60 


Water Powers of the Southern Appalachian System 35 

The Decker Cell 35 

Mica 36 


Illuminating Engineering 61 

Non-Sj nchronous Generators 61 

Standard Handbook for Electrical Engineers 62 

House Hill No. 10457 63 

The Central Station Distributing System Pressure Regu 

lation 64 

the Connecting and Repairing of Alternating 

• the Public Service Commission of the Light- 
ing Companies of New York 74 

Study Men ~ : 

Some Points about Series Transformers 77 

Commercial Day Program. N. E. L. A. Convention. 190S. 

Questions and Answers 79 

Electrolytic Copper Refining 

Downward Illumination 

Legal Ni >tes 

New Type of High-Si - am Engine 

JUN 12* 

5323 <* 




Large Electrical Machines Built at West Allis Works, 

Aliis-Chalmers Company 84 

Electric Equipment of Hydro- Electric Plant 85 

New Madeline Furnace of the Inland Steel Co., Indiana Har- 
bor, Ind 85 

Weston Portable Instruments 85 

Remarkable Performance of an Induction Motor 86 

General Electric Earns $9.800.000 86 

Joints 86 


Electricity from Coal 87 

The Delmar Short-Circuit Indicator 89 

New Patent Laws in Britain 89 

The Lighting of the Institute Library 89 

A Half Decade of Steam Turbine 89 

Central Station Distributing System 90 

Cost of a Single-Phase Line Equipment 96 

The Richmond and Chesapeake Bay Single-Phase Railway, 

Richmond, Va 97 

A Short-Circuit Interrupter 99 

The Problem of Illumination 100 

Downward Versus Horizontal Illumination 105 

Electric Locomotives — Continued 106 

Long Acre Company Hearing 109 

Legal Xotes no 

A New Type of Induction Motor in 

Aids to the Solution of Practical Illuminating Problems 111 

The Brist< <\ Company r 12 

A Xew Vertical Pump 1 1 _> 

Steam Turbine Sales 113 

Xew Allis-Chalmers Alternator for Xevada-California Power 

Company. Goldrield. Nevada 113 

Incandescent Lamps for Singer Building, Xew York 113 


Sapphire ami I )iamond Jewel- 117 

Meter Department of the Central Station [18 

Cable Insulation 127 

Compensators fur Measuring Line Drop 133 

Questions and Answers 134 

Lamp Testing 134 

General Electric Report 135 

Xew Breakdown Rate for Xew York 137 

American Circular Loom Co 137 

The Copper Handbook 1 38 

Large Railway Motor Contracts 138 

The General Electric Tungsten Lamp 13X 

Westinghouse Turbines for Manila and Japan 138 

Low Lighting Rates for Lafayette. Ind 138 

Williamsburgh Bridge Cables 138 


Returning Prosperity 139 

Westinghouse Readjustment is a Success 139 

The Transformer Station 140 

The Copper Situation 140 

The X. E. L. A. Gas Engine Report [40 

Motor for Steel Mills 140 

Testing Lamps by Substitution 141 

Some Points to be Considered in the Purchase of Steam 

Turbines 142 

The Central Station Distributing System 144 

Tape 151 

Receiving Stations Operated from High-Tension Transmis- 
sion Lines 153 

Distribution in Suburban Districts t6i 

Questions and Answers 162 

The Western Electric Company Enters the Steam Turbine 

Field 162 

Tlie National Electric Light Convention 103 


( reneral Electric Company's R< port 1 1 ; 

Downward vs. Horizontal Illumination 11; 

The Boron Jewel T r 6 


The QevelQpment of the Regulating Converter [65 

John A Roebling 


iv IN 


Motor Control System 165 

Kokomo, Marion & Western Traction Co 166 

Setting a Market Value on a Water- Power Plant 166 

Power Station Lighting 166 

Southern Water Power 166 

Protective Devices 167 

Printing Press Data 172 

1 .n »und Detectors -173 

Notes on Power Station Lighting 174 

The Synchronous Regulating Rotary Converter 175 

Power Required in Binderies 176 


The Electric Lighting Systejrn of the Washington Union Sta- 
tion, Washington, I). C 177 

Kokomo, Marion & Western Traction Co 182 

An Exhaust Steam Turbine Plant 187 


A Campaign 1 >t Education 191 

Transformer I ron 192 

Empire State Gas and Electrical Association 192 

Tiie Relation of Rates of Efficienc) of Light-Sources [92 

Artificial Lighting of Public Schools 103 

High-Tension Switchboard Practice in America 194 

A Remarkable Vacuum Pump 20!') 

Overhead Lines 207 

The Power Development of the Northern Colorado Power 

Companj 213 

Series Incandescent Systeftis with Tungsten Lamps 218 

A Tungsten Diffusing Cluster 210 

Trade Note's 22-1 

Questions and Answers 221 

Legal Notes 221 

Clippings from Consular and Trade Report^ 222 



Silicon Steel 22.-; 

An Old Controversy 225 

Electrical Shows : 

The New Haven Electrification 227 

Allis-Chalmers Annual Report 228 



Electric Exports 22S 

The Valuation of a Steam Power Plant 

Overhead Construction 229 

Impregnating Compounds ^5 

A New Business Problem 236 

Switchboard Notes 2 ^- 

E volution of the Return Circuit Department 238 

-Measurements with Portable Instruments 239 

New Plant for the Electric Cable Co 24; 

The New Westingftouse Lamp Works 246 


Permutators 247 

The We^tinghouse Reorganization 24S 

( Jutpul Costs of Small Plants 248 

l'h'. Business Outlook 249 

Safety Engine-Stops .".... 240 

Idle Permutator ; 2^0 

Electric Furnaces 254 

Currents Surge in Closing an Inductive Circuit 255 

Electrical Work in India ....... 257 

Underground Construction 2 ^ 

Comparative Cost of Various Street fllurhinants 263 

The Effibable Effect of the Higher Efficiency Incandescent 

Lamps on Central Station Income 264 

i he II. ctric Fault-Finder 269 

Idie Keokuk Accident 270 


Some Causes of Variations in High-Tension Practice 271 

Developments in Single-Phase Electrification 272 

Idle Return Circuit 273, 

< Mitpu' Costs of the Isolated Factory Plant 274 

Austin Municipal Plant 274 

Pennsylvania Terminal Electrification 274 

Some Features of European High-Tension Practice 275 

Tensile Strength of Trolley Wires 283 

Suburban Electric Railway Return Circuits 283 

On Protection from Lightning ' - 

Rushing a Telephone Switchboard to Paris 291 


FEBJZ Jfino 

Volume XXXIX. Number I. 
$1.00 a year; 15 cents a copy 

New York, January, 1 908 

Th e Electrical Age Co 
New York and London 

Published monthly by The Electrical Age Co., 45 E. 42d Street, New York. 

J. H. SMITH, Pres. C. A. HOPE, Sec. and Treas. 

Telephone No. 6498 38th. Private branch exchange connecting all departments. 
Cable Address — Revolvable, New York. 


United States and Mexico, SI. 00; 

Canada, SI. 50. To Other Countries, S2.50 


Insertion of new advertisements or changes of copy cannot be guaranteed for the following 
issue it received later than the 15th of each month. 


Business and Editorial Departments, 33 Bedford Street 

Strand. W. C. 

London Telephone. Post Office, 3807 Central. Private branch exchange. London Cable 

Address — Revolvable, London. 
All remittances to London Office should be crossed London and County Banking Company. 


Subscriptions may be sent to the following branch offices or agencies, where they will re- 
ceive the same careful attention as at the home office in New York : 

124, Queen St., Melbourne 33. Loop St., Cape Town 

359, George St., Sydney Unter den Linden, 5. Berlin 

61, Main St.. Yokohama Nevsky Prospect. St. Petersburg 

23, Esplanada Rd., Bombay 31, bis rue de Fauburg Montmartre, Paris 

General Agents for United States and Canada : The American News Company 

Report of the Public Service Com- 
mission, Second District, New 

In the first annual report of the 
Public Service Commission. Second 
District. New York, it is stated that 
the Commission will not consider ap- 
plications of foreign gas and electric 
corporations to do business in Xew 
York. In a general way. it is the 
policv of the State to foster and en- 
courage enterprises only where an en- 
terprise appears commercially feasible 
and likely to return a fair income upon 
the necessary capital. Strict adher- 
ence to this policy will, in great meas- 
ure, eliminate unprofitable enterprises 
from the money market and lessen 
very greatly the economical loss 
through the diversion of large 
amounts of capital in what would be 
otherwise unprofitable investments 
Strict adherence to this policv will 
prevent also destructive competition 
and obviate the multiplication of fa- 
cilities beyond the needs of the com- 
munity. Such a policy pursued by the 
State of Ohio would have prevented 
the condition of affairs now prevalent 
at Toledo, where neither of the op- 
erating central stations have been able 
to pay dividends upon their invest- 

The portion of the report dealing 
with the inspection of electric meters 
shows that the work of the Com- 
mission in this respect is farcical. 
Six tests of consumers' meters, upon 
complaint, have been made in six 
cases and the Commission stated, 
"after thorough investigation it is con- 
vinced that the tendency of electric 
meters is to under-register.'* though 
upon what grounds it arrives at this 
conclusion is not apparent. If such 
is the case, the central stations of New 
York are certainly entitled to ade- 
quate inspection in order to protect 
them against the losses from slow 

Power-factor Rating of Generators 

A large number of purchasers of 
electrical generators and other appara- 
tus for industrial plants do not fully 
appreciate the importance of power- 
factor. In general, if they buy, say. 
a 500 k.w. generator to care for ap- 
proximately 500 k.w. of induction mo- 
tors (with some allowance for line 
loss) they think they have linked up 

It is not generally clear to them 
that an induction motor generously 
large for a given duty may, by run- 
ning much underloaded for a large 
part of the time, raise the deuce with 
the output of their generator, not to 
speak of the exciter. The purchaser 
often makes two mistakes : first, by 
making a bad disposition of motors 
for a given service, or. in other words, 
a poor proportion of capacity to the 
average demand upon them : and. sec- 
ond, by buying the generator rated 
approximately at their /aggregate 
power for ioo per cent, power-factor, 
and the exciter on the same basis. 

Xow this is not particularly repre- 
hensible in many a purchaser, be- 
cause he really does not know any 
better. If he had an advising en- 
gineer this difficulty would not arise, 
but he does not. and often would not, 
because he thinks that the manufac- 
turer's engineer or salesman will see 
that he gets the proper apparatus. 

When it is considered how much 
free educational work in electrical en- 
gineering is done by the agents of 
electrical manufacturers simply 'to 
foster the use of alternating-current 
apparatus, it would hardly be ex- 
pected that any difficulty could arise 
over power-factor. But it does, and 
the situation usually develops as fol- 
lows : To sell a machine for, say, 
500 k.w.. 80 per cent, power-factor 
is mighty hard sledding against a 
quotation on one rated at 500 k.w.. 100 
per cent, power-factor, since the lat- 

ter is smaller and cheaper. The ex- 
citer quotations are likewise affected. 
Naturally, the purchaser is inclined 
to buy the cheaper machine since 
he well knows there is not much 
choice in the quality of apparatus. 
Even after the folly of purchasing 
a 100 per cent, power- factor rated 
machine for an 80 per cent, power- 
factor load is apparent, it is difficult 
to establish the real fact. For power- 
factor is an elusive thing, a sort of 
convenient bogy, which can be made 
to appear or disappear, and whose 
effects upon the generator may be 
attributed to heavy drop in the wiring 
to motors improperly sized to their 
work, or to numerous other condi- 
tions whose determination is expen- 
sive and whose cure is costly. For- 
tunately, the difficulty can be usually 
solved by the installation of a second 
machine. This remedy, however, is 
quite beside the question. 

This condition will probably not 
change until the matter is passed upon 
by the standardization committee of 
the Institute, and in the interest of 
the entire industry it is to be hoped 
that the committee will provide for 
a kilowatt rating at some power- 
factor common in industrial work. 
It may be best to approach *the mat- 
ter by defining the regulation of 
the machine. However, it is not for 
us to point out the proper solution 
of the difficulty. 

The Tungsten Filament 

One of the advantages of the 

tungsten lamp which is not gener- 
ally understood lies in the number of 
folds of the filament on itself. If the 
lamp burns out. or the filament breaks, 
it is possible to shake the lamp so that 
"iic of the broken ends will catch on 
to a good section of the filament. 
When the lamp is turned into the 
soeket again, the new filament con- 
tacts will fuse together, and the lamp 


January, 1908 

will continue to burn, though at an 
increased brilliancy due to the cutting 
out of a section of the filament. Re- 
cently a lamp which had burned out 
at 600 hours, was jiggered as de- 
scribed above and continued to burn 
300 hours longer. 

A New Storage Cell 

A new type of storage battery hav- 
ing, with the same ampere-hour ra- 
ting, about one-third the weight of a 
chloride accumulator, is being put on 
the market. The type of construc- 
tion is radical. By the enclosure of 
the lead oxid in a brittle and fragile 
earthenware box. made by placing two 
reticulated clay plates together, it is 
hoped to hold in the active material 
and to prevent the usual shrinking 
during charge and discharge, as well 
as to avoid the common enough buck- 
ling of the Plante plates. The cell is 
known as the "Standard" accumula- 

The active material and its lead- 
electrode are inclosed in a thin box 
of unglazed clay. The inner surface 
of the box is reticulated partly to 
strengthen the clay box and partly to 
grip the active material, while the 
outer surface is ribbed vertically to 
strengthen the clay plate further, and 
to permit the free circulation of the 

With a porous cell of this type it 
would be rather difficult to prevent 
infiltration of the active material and 
yet maintain a low internal resistance. 
This, it would appear, is altogether 
a matter of proper porosity, and to this 
problem. Mr. Clare, the inventor of 
the cell, formerly associated with the 
defunct Hatch Accumulator Co.. of 
Boston, which had infringed the pat- 
ents of the Electric Storage Battery 
Co., has brought twenty years of skill 
in clay working. It is believe by the 
inventor that this troublesome point 
has been overcome. 

The elemental clay box. which has 
just been described, is about 3 
inches square and half an inch thick, 
including the lead plate. By the as- 
sembly of these unit boxes accumu- 
lators can be made of any capacity. 
While the boxes are of quite fragile 
material, yet their compact form 
should make it possible to build a cell 
structure having greater strength than 
the individual clay box. After the as- 
sembly of the clay boxes, the whole 
is enclosed in glass plates nicely 
fitted together by grinding and ce- 
mented together by paraffine. The use 
of the enclosed glass cell with its 
paraffine cementation is an essential 
element in this accumulator. As 
closely as can be ascertained, none of 
the cells are as yet in central station 
or industrial service, and it will re- 

quire some time to establish beyond 
peradventure the durability of this 
type of construction. It is hardly 
likely to be employed in electric ve- 
hicles, owing to the jarring and jolt- 
ing to which the earthenware cells 
would be subjected. Xone of the 
gentlemen financially interested in the 
company have, so far as can be ascer- 
tained, any experience in the electric 
field, except the inventor himself. 

The Central Station Distributing 


The development of central-station 
distribution of electricity has been very 
rapid in recent years. Lower rates 
and increased advertising have 
brought about a condition, in some 
cases, which has required the best ef- 
forts of distributing engineers to keep 
up with the demands. 

The increasing number of men en- 
gaged in such work and the almost 
infinite variety of problems presented 
makes this an opportune time for a 
discussion of the principles govern- 
ing the design and extension of a dis- 
tributing system. 

The article in this issue entitled 
"Central-Station Distributing System" 
is the first of a series of articles cover- 
ing this subject which will appear in 
The Electrical Age during the suc- 
ceeding months. The principles of de- 
sign and the limitations met in prac- 
tice will be outlined concisely, and this 
will be followed by a discussion of the 
methods of construction, maintenance 
and operation which are current in 
American practice. 

These articles will endeavor to cover 
the field in such a way as to be of 
service to the engineers in charge of 
small distributing systems, as well as 
those who are directing the develop- 
ment of distributing systems in the 
larger cities. 


When you read the annual balance 
sheets of the great corporations of the 
country, do you imagine that the asset 
column really exhibits the most valu- 
able and best monev-makins: elements 
of the concern ? Do the figures rep- 
resenting the value of the plant, the 
patents, the working capital or the 
volume of business done annually give 
a true conception of the actual pros- 
perity of a concern? Is it possible 
from these figures to form a sound 
judgment as to whether the stock of a 
company ought to go up or down in 
price? Surely not. Truthful as may 
be the figures representing the money 
value of a business, they fail totally 
in giving an idea of the most impor- 
tant element necessary for the success 
of an industrial plant. 

Personality is of more consequence 
to success than all the elements shown 
in the balance-sheet figures put to- 

One dominant man who absolutely 
commands the respect of all who come 
into professional contact with him is 
a tower of strength to even the 
largest corporation, and with a corps 
of such men at the heads of depart- 
ments any corporation can defy com- 

Perhaps no more remarkable in- 
stance of this tremendous influence 
can be cited than that of Edward 
Reynolds in the sale of the 96,000 
h.p. engine equipment of the Manhat- 
tan power-station. Xew York. Could 
any other man in the world have sold 
96,000 h.p. of engines of a size and 
type never before built upon the 
strength of nothing but his reputation 
and a plan hastily sketched upon the 
back of a visiting card? 

That was what Mr. Reynolds did. 

The Manhattan Railway Co. had 
arranged to replace its old steam loco- 
motives by electric traction. The 
company had secured for its power- 
house the only available water site 
within the proper distance of its lines 
and figured that it needed 96,000 
nominal horse power with perhaps a 
50 per cent, overload capacity for 
-hort-load peaks. 

There was no question about how- 
much space would be absolutely re- 
quired for boilers, and when this was 
figured out it was clear that if the 
necessarv horse power in engines and 
generators were to be installed it 
would have to be done in a space much 
smaller than any such amount of 
prime movers had ever been gotten 
into before. 

Many engineers of repute had been 
consulted about the design of the en- 
gines, but none had offered a satis- 
factory solution of the problem. 
Finally' Mr. Reynolds was sent for. 
The summons was by wire, asking him 
to meet the board of directors in New 
York the following day. 

Mr. Reynolds was in Milwaukee 
when the summons came. He started 
for Xew York at once. As a designer 
and builder of big steam engines Mr. 
Reynolds was known all over the 
world. Under his supervision the big 
engine of the Centennial Exposition 
in Philadelphia, in 1876, had been built 
at the Corliss works, where Mr. 
Reynolds worked his way up from 
the bench to the place of general man- 
ager. As chief engineer of the E. P. 
Allis works he had designed and built 
the big engine of the Columbia Ex- 
position at Chicago in 1894. What- 
ever he took in hand he handled suc- 
cessfully. His word was a guaran- 

Mr. Revnolds started from Mil- 

January, 1908 


waukee. The story of the inception 
of the Manhattan type of engine, as 
he told it, is as follows : 

"As I rode from Milwaukee to 
Chicago, I began to consider how to 
get 96,000 h.p. of engines and genera- 
tors into the required space. I took 
a mental survey of every type of en- 
gine that had ever been built and 
considered them one by one. I saw 
that it was possible to use one of the 
accepted types of engines for the pur- 
pose, but in order to secure the con- 
stant turning moment necessary for 
the close regulation of electrical gen- 
erators of this type it would require 
fly-wheels so heavy that no concern in 

America could build shafts big enough 
to carry them. 

"I had arrived at this conclusion 
when my train reached Chicago and 
I transferred to the Xew York train 
and again took the subject under con- 

"I thought and thought about it 
until bed time, and finally turned into 
my berth without having arrived at 
any solution. 

"As I lay in my berth there sud- 
denly flashed into my mind a way to 
overcome the difficulty. I turned on 
the light in my berth, fished out a 
visiting card from my pocket, and 
with a lead pencil made a hasty sketch. 

Then \ turned over and went to sleep. 

"With the pencil sketch to show 
what I proposed to build I went into 
the meeting of the board of directors 
the next day and took the order for 
the engines." 

Mr. Reynolds built the Manhattan 
engines according to his original in- 
spiration. They not only worked, but' 
although the power-house was short 
two engines during the whole of the 
first year of its use it never had to shut 
down nor did the railroad company 
ever have to take off a train or restrict 
the service even during the most 
crowded rush hours. 

The Estimation of Power-Factor on Unbalanced 

Three-phase Loads 


THE following method has been 
found useful in the estimation 
of power-factor on unbalanced 
three-phase loads and also in check- 
ing series meter connections where 
other means are found unavailable. 

Before going into explanations, a 
case will be worked out from read- 
ings actually taken. 

Figure 1 shows wiring of feeder in 
hand. Voltages (a) to {b), (b) to 
(c) to (a) are found to be alike, 
hence the voltage vectors (a) to (b), 
(b) to (c) and (c) to (a) form an 
equilateral triangle. Had (a) to (c) 
measured 73 per cent, greater than 
(a) to (b), then there would have 
been a cross in. the shunt connections 

The next step is to determine which 
e.m.f. is leading; (b) to (a), or (b) 
to (c). 

Three Phase 120C0 Volts 

FIG. I. 

This determination is in most casej 
unnecessary because the complete 
vector diagram can be worked out 
first and it will then be apparent which 
rotation of diagram suits the nature 
of the load in question. 

For this determination a Weston 
wattmeter (or any wattmeter whose 
shunt current is in step with its shunt 
e.m.f.), an incandescent lamp and a 

spare wattmeter whose shunt circuit 
forms a good choke coil are used. 
The lamp is placed in series with the 
series coil of the wattmeter (see Fig. 
2). The choke coil is placed in series 
with its shunt. The remaining ends 
of shunt and series are connected to 
shunt lead (b). 

Shunt Lead 11 

— *- 





Shunt Lead /. 

O ( 


Shunt Lead C 

— fr- 





\\ atti 


FIG. 2. 

Then (x) is connected to (a) and 
(y) to (c) and the wattmeter read- 
ing is 22 watts. Then (x) is con- 
nected to (c) and (y) to (a) and the 
reading is 69 watts. This shows that 
the e.m.f. (b) to (a) lags behind the 
e.m.f. (b) to (c) because on the first 
reading of 22 watts the lamp load, 
being ohmic, gives the current to the 
series of "wattmeter in step with e.m.f. 
(b) to (c), and the e.m.f. (b) to (a) 
lags that by 60 degrees, and the 
choke coil makes the shunt current 
lag still more, giving a very low 
power-factor leading current applied 
to the wattmeter, whereas on the 
second reading of 69 watts, the series 
current is in step with e.m.f. (b) to 
(a), and the e.m.f. (b) to (c) leads 
it by 60 degrees, and the choke coil 
makes the shunt current lag behind 
this 60-degree lead, giving a compar- 
atively high power-factor applied to 
the wattmeter, and hence the higher 

reading of the meter for the same 
volt-amperes on meter. 

Xow that we know that the e.m.f. 

(b) to (a) lags behind the e.m.f. (b) 
to (c) we can plot the e.m.f. triangle 
(a), (b), (c), accordingly as in Fig. 
3, and the order in which the e.m.f.'s 
over neutral reach maximum is (a), 
(b), (c). 

Returning to polyphase wattmeter 
Fig. 1, the load current is following 
in its series. Connect shunt lead (a) 
to shunt coil (a) and leave shunt coil 

(c) dead. The reading of meter is 
500 kw. Now connect shunt lead (a) 

fig. 3. 

to shunt coil (c) and leave shunt coil 
(a) dead. The reading is 400 kw. 
Connect shunt lead (c) to shunt coil 
(c) and leave shunt coil (a) dead. 
The reading is 700 kw. Connect 
shunt lead (c) to shunt coil (</) and 
leave shunt coil (c) dead and the 
reading is minus 250 kw. 

Construction. — In Fig. 3, starting 
from center of triangle measure a 


January, 1908 

distance parallel to direction of (ba), 
of 500 kw., taking scale at 600 k\v. 
equal one inch. From this point (h) 
draw a line (h i) perpendicular to 
direction of (ba). Again starting 
from center of triangle measure par- 
allel to direction of (be) a distance 
of minus 250 kw. That is to say, 
measure 250 six-hundredths of an 
inch in direction parallel to (cb), as 
shown to point (k). Draw (k i) 
perpendicular to (be), and where 
(k i) intersects < // 1) is the end of 
the current vector of the current in 
(A) wire of feeder. Similarly meas- 
ure from center parallel to (be) 700 
kw., and erect perpendicular. Meas- 
ure 400 kw. from center parallel to 
(b a) and erect perpendicular. The 
intersection of these perpendiculars is 
the end point of current vector of cur- 
rent in (C) wire of feeder. 

Now, if circuit is three-wire, three- 
phase, then complete the parallelo- 
gram of which the two current vec- 
tors are the two sides. Extend its 
diagonal through the center to an 
equal distance on other side of center, 
and the end of this is the end of the 
third current vector. 

The lengths of the current vectors 
so plotted measure the actual am- 
peres in the three wires, and the 
directions of these vectors relatively 
to lines drawn from the center of the 
triangle through its corners (a), (b) 
and (c) respectively show the angles 
of lag of the currents in the three cor- 
responding wires. 

Methods have been published for 
obtaining power-factor on balanced 
three-phase loads, by ratio of watt- 
meter readings, but these are ren- 
dered quite deceptive when used on 
unbalanced loads. 

If the e.m.f. between wires i>. say, 
12,000 volts, then the scale of am- 
peres on which the current vectors 
must be measured is 600,000/12,000 
= 50 amperes per inch when kilowatt 
scale is 600 kw. per inch, or 600,000 
watts per inch. 

The correctness of the above 
method can be easily proved by show- 
ing that the current vector A is the 
only line whose component along the 
e.m.f. (b) to (a) is plus 500 kw.. 
and at the same time whose compo- 
nent along (b) to (c) is minus 250 
kw. Similarlv with current vector 


Xow from the above kilowatt read- 
ings the total real kilowatt of the load 
is 500+700=1200 kw. 

Applying these values to the old 
formula to give power-factor by ratio 
of wattmeter readings we have : 

1+500 700 

2 V ( 500 '700 ) - — 500 700+ 1 
= . 9 G. 

This corresponds to 16 degrees lag. 

The- diagram Fig. 3 shows the cur- 
rent in A to be 64 amperes, and to 
have a lag of 20 degrees ; the current 
in B to be 75 amperes, and to have a 
lag of 30 degrees ; the current in C to 
be 59 amperes and to have a lag of 
35 degrees. 

A method used for calculating 
power-factor in unbalanced three- 
phase circuits by a large electric man- 
ufacturing company is to divide the 
total real watts by the volts across 
lines times the average amperes per 
wire, times 1.73. This method ap- 
plied to the above amperes and watts 
gives a power-factor of 87.3 per cent., 
which corresponds to a lag of 29 

The American Institute of Elec- 
trical Engineers gives a definition of 
power-factor as "ratio of real watts 
to the apparent watts." The appar- 
ent watts are the product of the volts 
and the amperes, when applied to a 
single-phase circuit. When applied 
to a two-phase circuit with balanced 
load, the amperes, or. more correctly 
speaking, the equivalent single-phase 
amperes, are twice the current in one 
wire, while the volts are the volts 
measured across one phase, or across 
opposite corners of the voltage square. 
In a three-phase circuit with balanced 
load, the equivalent single-phase am- 
peres are 1.73 times the amperes in 
one wire, while the voltage is the 
voltage from wire to wire, or equal to 
one side of the e.m.f. triangle. 

Xow when we come to unbalanced 
three-phase loads the power-factor 
has been defined as the ratio of the 
total real watts to the apparent watts 
or (volt-amperes) where the am- 
peres are 1.73 times the average cur- 
rent per wire, and the volts are the 
volts between wires. 

Since it is the net wattless am- 
peres that is the undesirable feature 

of low power-factor, and as these 
wattless amperes are a very real 
quantity rather than an apparent 
quantity, why not define the power- 
factor in unbalanced three-phase cir- 
cuits, or indeed in any circuit, by say- 
the angle of lag or lead is the angle 
whose tangent is the algebraic sum 
of the wattless amperes divided by the 
algebraic sum of the working com- 
ponents of the currents (taking lag- 
ging wattless amperes as, say, plus and 
leading wattless amperes as minus). 

It can be shown by geometry that 
the formula to give tangent of angle 
of lag (for balanced or unbalanced 
three-phase circuits) according to the 
above definition is : 

tan <l> = 

(W a -W.)- 2 (W ac - 
V 3 "(W a -r-W ) 

W ca ) 

Where W a equals reading, using 
current A reacting on e.m.f. (b to a). 

W c equals reading, using current C 
reacting on e.m.f (b to c). 

W ac equals reading, using current A 
reacting on e.m.f. (b to c). 

\Y ra equals reading, using current C 
reacting on e.m.f. (b to a). 

In the above case 

\Y a = +500 kw. 
W c = +700 kw. 
\\\ c = — 25okw. 
YV ca = +400 kw. 

Tan <l>=+o.53 
*=28° lag. 
l'F.— 88.3 per cent. lag. 

This checks pretty well with previ- 
ous method and requires only watt- 
meter readings, which are much more 
reliable than alternating-current am- 
meter readings, especially on light 
loads. On specially unbalanced loads 
this formula gives much more reason- 
able results. Take the case of an in- 
candescent lighting load across one 
phase of a three-wire three-phase cir- 
cuit. This formula gives 100 per 
cent, power-factor while the previous 
method gives 86.6 per cent. As a 
further illustration, take similar lamp 
loads across two of the three phases 
we have 100 per cent, power-factor 
by the proposed formula, and the 
previous method gives 93 per cent. 

The Central Station Distributing System 

Feeder Calculations 


General Inspector, Commonwealth Edison Co. 

IN the transmission of electricity 
over wires, energy is lost by dis- 
sipation as heat, in proportion to 
the square of the current and to the 
resistance of the circuit. 

Loss=C 2 R watts, 
when C is in amperes and R in ohms. 

As the current in a circuit carrying 
a given amount of power decreases 
with increase of working voltage, it 
is obvious that the designer has con- 
trol of the energy loss on his circuits 
in two ways, viz., by varying the volt- 
age at which current is delivered, and 
by varying the resistance of the cir- 

The voltage at which current is 
delivered to the patrons of a distrib- 
uting system must, on account of risk 
to life and property, be comparatively 
low., i.e., ioo to 250 volts. Current is 
delivered for power purposes at 500 
volts, and in some cases at 2200 volts 
for large consumers, but these volt- 
ages are unsuitable for lighting and 
small power purposes. 

These voltages being fixed by other 
considerations, the problem resolves 
itself into one of selecting the proper 
resistance., i.e., the proper size of con- 
ductor for the load. 

Copper being obtainable in quan- 
tities and having a low specific resist- 
ance, is used as a conductor for dis- 
tributing systems exclusively. Alum- 
inum has been used on transmission 
lines for some years, but is not suit- 
able for general purposes, owing to 
the difficulty of making joints and 

The resistance of copper as deter- 
mined by Mathiessen and adopted by 
the American Institute of Electrical 
Engineers in 1893 is 9.59 ohms per 
mil-foot at o°C. That is, a wire of 
pure copper 1 ft. long and 1/1000 
inch in diameter has a resistance of 
9.59 ohms at the freezing point. Com- 
mercial copper is usually about 98 per 
cent, pure and for wire as used, the 
resistance per mil-foot is therefore 
about 9.68 ohms. 

The resistance of copper increases 
as the temperature increases, at the 
rate indicated by the following for- 
mula : 

R= 9 .69 ( 1 +.00406 T) 

in which T is the temperature in de- 
grees centigrade. 

At 1 5 Cent, (equivalent to 59 

Fahr.) R=9-68 [1 + (.00406 by 15)] 
= 10.27. 

This represents a temperature com- 
mon in distributing work overhead 
and underground in the temperate 

The resistance of one mil-foot at 
various temperature is given in the fol- 
lowing table : 

To illustrate : 100 amperes is to be 
transmitted 200 feet at a drop of two 
volts, what size wire will be needed? 

eg. C 

Deg. F. 

R. per mil- 


— 4 






























11 .24 













The resistance of one foot of wire 
of one circular mil area being 10.27 at 
59 degrees Fahr., the resistance of D 
feet would be 10.27 by D. Likewise 
the resistance of one foot of wire hav- 
ing M circular mils area would be 
to. 27 'M. 

Therefore the resistance of a wire 
D feet long of M circular mil. area 
is R=D x 10.27 ^ r - 

By the law of Ohm the drop in 
voltage in a circuit carrying C am- 
peres and having a resistance of R 
ohms is E=C R. 

Since for copper wire R=D by 
10.27 M the drop in voltage of D feet 
in copper wire of M circular mils area 
is E=CD x 10.27 AI. 

If D is taken as the distance one 
way, the drop in a two-wire direct- 
current circuit is : 


2 DC x 10.27 DC x 20.54 



Or if it is desired to know what size 
wire will be required to carry a load 
of C amperes a distance of D feet at 
a drop of E volts 



-=205,400 C mils. 


DC x 20.54 

By reference to the wire table it is 
found that this is nearly the area of 
0000 B. & S. gauge wire, and this 
size would therefore be selected. 

In many cases the problem consists 
in ascertaining what the drop will be 
in a certain feeder whose length and 
size are fixed. In such cases the re- 
sistance per 1000 feet may be used to 
good advantage as follows : 

E=C x D x R 
in which D is the number of thou- 
sands of feet and R the resistance per 
1000 feet, or 10.27x1000 /M. 

A feeder of Xo. o. 5000 feet long, 
is to carry 80 amperes. What will 
be the drop? The value of R for Xo. 
o wire is : 

10.27 x 1000 

= -0974 ohm. 


From the rule above E=8o x 5 x 
.0974=38.9 volts. 

The value of resistance per 1000 
feet of the various sizes of wire may 
be found in Table I. 

When alternating current is trans- 
mitted, the foregoing rule for deter- 
mining the drop in voltage is not ap- 
plicable, since the variations of the 
magnetic field accompanying the al- 
ternations of the current produce 
"back pressure" in the circuit which 
exerts a choking effect on the im- 
pressed electromotive force. This is 
called the electromotive force of self- 
induction, or the inductive reactance. 
It varies with the size of wire and the 
distance between centers of the wire. 
being less for small separations and 
more for greater ones. 

The drop in pressure in an alter- 
nating circuit is therefore composed 
of two elements: (a) the drop due to 
the resistance of the wire, which is 
proportional to the energy loss and is 
known as the ohmic drop, and (b) the 
drop due to the self-induction of the 
wires acting as a closed circuit, known 
as the inductive drop. 

The ohmic drop is that which would 
occur in the circuit if it were carrying 
a direct current of the same volume. 


January, 1908 

The inductive drop, which is a 
quarter cycle behind the current, do^ 
not subtract directly from the im- 
pressed voltage as does the ohmic 
drop in a direct-current circuit. 

These two components of voltage 
drop may be represented as the two 

atus. which is represented by AB in 
Fig. i. 

The flow of alternating current in 
any piece of apparatus composed of 
coils having an appreciable inductive 
reactance is therefore governed by the 
ohmic resistance and the inductive 





Per 1,000 Feet 





Bare Braid 
W. P. 








12.4 25 
19.8 35 
31.4 53 





.128 .147 
.162 .180 
.182 i .209 

50.0 74 

79 5 111 

100.2 135 




52 634 





. 263 












253.4 310 
319.7 407 
403.0 495 




.410 1 .475 
.460 | 524 










576 920 

908 1,041 

1.059 1.231 








1 362 
1 514 

1 .420 

















1 . 224 

1 . 296 
1 652 





sides of a right-angled triangle of 
which the third side is the resultant. 

The feeder circuit forms a closed 
loop through which all the lines of 
force of the magnetic field are linked. 
In a similar manner, the windings of 
motors, arc lamp coils, solenoids are 
linked with the magnetic field set up 
by their current flow. The field of 
force in such apparatus is, however, 
many times stronger than that of the 
feeder circuit, owing to the use of iron 
cores which have a very low resist- 
ance to the passage of magnetic flux. 
The use of a considerable number of 
turns in the winding, combined with 
the stronger magnetic field with a 
given current strength, causes all such 
apparatus to have a high inductive 

When an alternating voltage is ap- 
plied to such devices the flow of cur- 
rent, at the instant the circuit is 
closed, is proportional to the resist- 
ance of its windings and the pressure 
applied. As the pressure is changing 
from instant to instant, the current 
tends to follow the variation of the 
pressure. This change of current 
strength results in a change in the 
strength of the magnetic field, and 
this in turn induces an electromotive 
force in the windings of the appar- 

reactanee, and in the case of mot 
transformers and arc lamps by the 
"counter electromotive force, which 
represents the useful work being per- 
formed by the apparatus. 

The inductive reactance set up by 
the current which flows under the in- 
fluence of the impressed pressure is 
not directly opposed to the impres 

FIG. I. 

pressure as are the resistance and 
counter electromotive force, but is 
represented by the line AB at right 
angles to OA, the electromotive force 
doing work. 

The impressed pressure on the ap- 
paratus is OB, and the current in the 
apparatus is out of phase with the im- 
pressed pressure by the angle AOB. 
The power in the apparatus is propor- 
tional to the product of the current by 
OA and not by OB. as it would be in 
a non-inductive circuit. 

A wattmeter may be used to meas- 
ure the power put into the circuit 

directly in watts. The product of cur- 
rent by voltage OB is known as the 
apparent power and is expressed in 
"volt-amperes. T ' The ratio of the 
actual power to the apparent power is 
called the '"power-factor." The volt- 
amperes put into a circuit may there- 
fore be multiplied by the power-factor 
tc obtain the actual power, where no 
wattmeter is available. The power- 
factor of induction motors varies from 
50 per cent, or less at starting up to 
85 per cent, or 90 per cent, at full 
load. Large motors have a higher 
power- factor than smaller ones of the 
same type. 

In Fig. I, OA being the power com- 
ponent of OB. the ratio of OA to OB 
is the "power-factor." Likewise AB 
being the inductive component, the 
ratio of AB to OB is known as the 
"inductance factor." The power- 
factor being cosine of the angle AOB, 
the inductance factor is the sine of 
this angle, and the inductance factor 
may thus be readily found by refer- 
ence to a table of sines and cosines, 
if the power-factor is known. 

In an electric light feeder the com- 
ponent of the drop due to the resist- 
ance of the conductor may be repre- 
sented by ( )A. and the component of 
drop due to inductance by AB. In 
an overhead circuit AB varies as the 
distance between wires is varied. In 
underground cables the conductors are 
usually so close together that the 
effect of inductance is about half that 
in overhead wires strung from 12 to 
18 inches apart. 

The resultant OB for a given 
feeder is, however, not necessarily the 
net drop in voltage impressed at the 
far end of the feeder. This varies 
with the power factor of the load 
though the same current is carried on 
the feeder in each case. That is, if a 
certain load draws 100 amperes at 70 
per cent, power-factor, and another 
load draws 100 amperes at 100 per 
cent, power-factor, at the same deliv- 
ered pressure, the net drop will be 
greater with the 70 per cent, power- 
factor current than with the 100 per 
cent, power-factor current. 

In Fig. 1, OA represents the ohmic 
and AB the inductive drop in a cir- 
cuit of Xo. o wire strung 12 inches 
apart. Under this condition the ohmic 
component is nearly equal to the in- 
ductive. With wires larger than X". 
o the inductive component is greater 
than the ohmic. With smaller wires 
the ohmic component is the greater. 
The relative values for Xo. 6 wire are 
represented by OA and CD in Fig. 1. 

Referring to Fig. 2, let the line OE 
represent the pressure delivered at the 
terminals of an induction motor. 
OR is the component of OE. which is 
doing useful work. ER is the watt- 
less component of self-induction 

January, 1908 


which causes the current in the motor 
to be out of phase with the impressed 

EL is the ohmic loss in the line, and 
LP is the inductive component of the 
line drop. The ohmic component of 
the line drop EL and the power com- 
ponent of the impressed voltage are 
in phase and therefore add directly. 
Similarly ER and LP are added. The 

resultant OP is the bus pressure 
necessary to deliver a pressure OE at 
the motor terminals. The drop is 
therefore the difference between OP 
and OE. 

With non-inductive load, such as 
incandescent lamps, ER disappears 
and the impressed pressure on the 
lamps takes the position OF (=OE). 
The generated pressure necessary to 













1 Degrees 

P. F 





76 6 






Per cent. 

Ind. F 











Per cent. 


By means of the table calculate the Resistance- Volts and the Reactance- 
Volts in the line, and find what per cent each is of the e.m.f. delivered 
at the end of the line. Starting from the point on the chart where the 
vertical line corresponding with power-factor of the load intersects the 
smallest circle, lay off in per cent the resistance e.m.f. horizontally and 
to the right ; from the point thus obtained lay off upward in per cent the 
reactance-e.m.f. The circle on which the last point falls gives the drop 
in per cent of the e.m.f. delivered at the end of the line. Every tenth 
circle arc is marked with the per cent drop to which it corresponds. 



■•r. >-f <! 

Throughout the table the lower figures give 
values for one mile of line corresponding to 
those of the upper figures for 1000 feet of line. 






a ~ 



«? V 

u ft 

& ° c 

U «M 


3 _ V 



V - C 


Upper figures are Reactance-Volts in 1000 ft. 
of Line (^2000 ft. of Wire) for One Ampere at 
7200 Alternations per Minute (60 Cycles per Sec- 
ond) for the distance given between Centers of 







9" |12"' 

i j 

18" 24" 




639 ; .098 .046 





.180 ! . 193 

.212 '225 















1 .26 













































.475 .639 



























1 .10 














































1 .65 

.07 '«) 

. 560 














1 59 






















. 1 99 













1 .05 



















































































































deliver OF at the lamps is ON and the 
drop is the difference between ON 
and OF. 

The power-factor being the cosine 
of the angle EOR, and the inductance 
factor being the sine of the same 

B i' 

FIG. 2. 

angle, the inductance factor may be 
readily found in any case when the 
power-factor is known by reference 
to a table of sines and cosines. For 
convenient reference such a table of 
corresponding power and induction 
factors is provided herewith in Table 
II. The reactance and resistance of 
the ordinary sizes of wire, at various 
separations, are contained in Table 

To illustrate : assume an induction 
motor load of ioo amperes at 2200 
volts single-phase delivered at the end 
of a line of No. o wire 4500 feet long 
with wires 12 inches apart, a fre- 
quency of 60 and a power-factor of 
80 per cent. The power-factor of the 
load being 80 per cent., we find by 
reference to Table II that the induct- 
ance factor is 60 per cent. 
OR is therefore .80x2200=1760 volts. 
ER is therefore .6 x2200=i320 volts, 

By reference to Table III, we find 
that the ohmic loss per 1000 feet per 
ampere for No. o is .2 ohm. Hence 
EL is .2 x 4.5 x 100=90 volts. 

Similarly, the indvictive drop per 
1000 feet per ampere for 12-inch 
centers being .22, LP is .22x4.5x100 
=99 volts per wire. 

OR+EL is 1760+90=1850 volts. 
ER+LP is 1320+99=1419 volts. 

The resultant of these, OP, is 
V( 1850) 2 + ( 1419) 2 =2332 volts. 

This is the pressure necessary to 
deliver 2200 volts at the end of the 
line. The drop is therefore the differ- 
ence or 132 volts, with a load of 2200 
X 1 00 X -8 =176,000 watts. 

If a lighting load of 100 amperes at 
100 per cent, power-factor were being 
carried, the inductance factor ER 
becomes zero and ON is 

Table for Calculating the Drop i.. Alternating-Current Liner 

V(229o) 2 +(99) 2 =2292 volts. 

The drop is therefore 92 volts, with 
a load of 2200X10QX 1=220,000 



January, 1908 

If the line were a three-phase three- 
wire line, carrying ioo amperes per 
wire of motor load, the values of EL 
and LP would be multiplied by .866. 

The bus pressure would be 
OP= V( 1838) 2 + ( 1406)^=2314 volts. 

The drop would therefore be 114 
volts instead of 132, as on the single- 
phase motor line, with a load of 
3 X 2200 X 100 X -577 X .8=304,500 

In a four-wire three-phase system 
with 2200 volts from phase to neutral, 
the drop is one-half that on a single- 
phase feeder carrying the same cur- 

Mershon has devised a diagram by 
which the long calculations of the 
pressures involved in problems which 
are to be solved by the method of Fig. 
2 are avoided and the result obtained 
with very little effort and with suffi- 
cient accuracy for all ordinary pur- 

This diagram is presented in Fig. 
3. The concentric circles are drawn 
from a center off the page which cor- 
responds to the point o in Fig. 2. 
In order to make the diagram appli- 
cable to any voltage, the rectangular 
divisions are divided into percentages. 
In the case used as an example in the 
foregoing the drop would be deter- 
mined thus : 

From Table III the ohmic loss 90 
volts, 4.1 per cent., and the inductive 
loss 99.0 volts, 4.5 per cent., is first 
ascertained. The intersection of the 
.8 power-factor ordinate and the inner 
circle corresponds to the point R in 
Fig. 2. Beginning at this point pass 
to the right four divisions, and thence 
vertically upward 4^2 divisions. The 
point thus reached corresponds to the 
point P in Fig. 2. The circle which 
passes nearest this point is the six per 
cent, circle. The drop is therefore 6.0 
per cent, of 2200 volts, or 132 volts. 

Each small division 
equals one percent 

.7 .8 


O «0 20 30 


Chart for Calculating the Drop in Alternating Current Lines 

fig. 3. 

Modern Arc Lighting 


THE arc light, having once es- 
tablished its supremacy as a 
street illuminant, made little 
progress for many years. The inven- 
tion of the Welsbach gas mantle, by 
restoring to the gas light much of its 
old-time prestige, has given an im- 
pulse to the arc light which has not 
only made it overtake the gas mantle, 
but has placed it in a position of 
greater supremacy than ever. 

The theory of the arc light has 
similarly undergone a transformation 
which may well surprise those who 
have not closely followed its evolu- 

An account of modern ideas and 
modern practice is given below : 


An electric arc is the passage of 
current between two terminals 
through a conducting vapor bridge 
consisting of the material of the nega- 
tive terminal, issuing from the nega- 

The electric spark, on the other 
hand, is merely the passage of cur- 
rent between terminals through the 
medium of the gas or vapor filling 
the space. 


The spark will pass as soon as 
there is a sufficient voltage,because the 
conducting medium is there. The arc 
cannot pass without first establishing 
the bridge of conducting vapor from 
the negative to the positive. This re- 
quires the expenditure of energy for 
the latent heat of evaporation, kinetic 
energy of motion of the vapor stream. 
etc., and this energy must be ex- 
pended before the arc can exist. Be- 
fore the current flows, no energy can 
be expended by the electric circuit, 
and therefore it follows that the elec- 
tric arc by the law of conservation of 
energy cannot spontaneously start, 
but must be started. 


( 1 ) Bring the conductors in metal- 
lic contact with each other, and then, 
the circuit being established, separate 
them, the current flowing and the en- 
ergy required to produce the vapor 
bridge being derived from the elec- 
trical energy of the current. This 
method is used in arc lamps. 

(2) Raise the potential across the 
terminals so high that the energy of 
the electrostatic field between the 

terminals is sufficient to produce the 
vapor bridge. 

(3) Supply the vapor bridge from 
a subsidiary arc. This is made use 
of in the mercury arc rectifier. 


Until very recently the only arc 
used for illumination was the arc be- 
tween carbon electrodes, though at 
the present time there are several ma- 
terials used, such as magnetite, mer- 
cury, and mixtures of various sub- 
stances with carbon. 


The vapor column of the carbon 
arc impinging on the positive elec- 
trode raises the end of the latter to 
a very high temperature and volatil- 
izes the carbon. The effect of this 
is to cause a hollow to be burned in the 
tip of the positive electrode. This 
hollow is called the crater. 


The light of the arc is derived from 
the vapor column or from a body 
heated to incandescence by the vapor 

In the old carbon electrode arc no 
special effort was made to raise the 
luminosity of the vapor column, the 
incandescent spot or crater on the 
positive carbon being relied upon for 
practically the entire light. For this 
reason the positive carbon was always 
placed above the negative in order 
that the crater might better shed its 
light downward. 

It is not improbable that in the near 
future lamps will be developed with 
luminous vapor columns impinging on 
positive electrodes of high luminosity. 


The following tables give data re- 
garding arc lamps of various types 
for direct and alternating current 
circuits. Among the important items 
are the watts per candle-power at no 
volts, and the length of carbons used 
per hour. These lamps are compar- 
able in efficiency of distributing the 
electric energy, the watts per candle- 
power at no volts being a true meas- 
ure of the lamp efficiency. The column 
headed carbons used per hour is of 
value, since for any allowable length 
of carbon the interval between trims 
is definitely fixed. 


When the crater is surrounded by 
the hot gases constituting the arc, the 
arc is silent, but if the crater flows 
over the tip of the carbon and comes 
into contact with cool air, a peculiar 
hissing sound is evolved. 


A carbon arc enclosed in a glass 
globe runs much longer than an open 
arc, owing to the slower consumption 
of the carbons. The color of the en- 
closed arc is usually pale violet and is 
disliked by the public. 

While it is possible to run an en- 
closed arc so as to give a good color, 
it is usually found desirable to 


Type of Lamp. 



Useful, j Total. 

cal in- 

per candle- 






Ordinary Carbons 




51 6 











241 2 




_, s _, 











4 ,814 

4 800 


1 .339 













2 334 

. !>fi2 








Intensive flame arc. inclined cored carbons.. . 
Enclosed arc (American) 


Magnetite arc 

Bremer Lamp (9-amp.) 

Carbo-mineral lamp (9-amp.) 

Carbo-mineral lamp (5-amp.) 

Carbo-mineral lamp (3-amp.) 



Carbone Luminous arc 


The flame arc, unlike the arc de- 
scribed above, owes the greater part 
of its luminosity to the vapor column 
and very little or none to the heat of 
the positive electrode. 

lengthen the arc, so that it will run 
directly off the no or 125 volt mains. 
It is the abnormally long arc which 
has the disagreeable violet color. 
Open arcs cannot be run at high volt- 




January, 1908 

age, owing to the crater on the posi- 
tive carbon overflowing and being 
cooled by the air when the voltage 
is raised considerably. For this reason 

the negative carbon, being wide and 
flat, cuts off much of the light from 
the crater, and makes this type of arc 
less efficient than the open arc. 


Type of Lamp. 

Ordinary carbons 

Ordinary cored carbons 

Flame arc, vertical carbons .... 
Flame arc, inclined carbons.. . . 

Enclosed arc 

Bremer lamp 

Carbo-mineral lamp 

Carbo-mineral lamp (3 in series) 




















Useful Total 






spheri- | power 


*$*?- Abso 


























[Both of above tables from Blondell, Soc. Int. Elect. Bull. 7, pp. 137-169, March, and pp. 267-286, April, 1907. 


open arcs cannot have the violet color 
characteristic of a long arc. 


A series of tests made with different 
clearances between carbons and with 
different cap openings show the total 
consumption of carbons to vary with 
the air space. 







of top 






per hour 

per hour 































This is best shown by a series of 
tests made on a well-known make of 
enclosed arc lamp in which the photo- 
metric measurements were made at 
different angles from the horizontal. 

The candle-power (C. P.) was 
taken by Ayrton's method, using red 
and green glasses separately to ana- 
lyze the light. 


A peculiarity of the solid carbon arc 
is that with any particular length of 
arc, if the current be increased, the 
difference of potentials across the car- 
bons will be decreased. 

This occurs continuously until a 


The consumption of carbon is least 
when the arc is near the top of the 
glass, i.e., near the opening, and great- 
est when the arc is at the bottom of 
the glass. 


The current being small and the arc 
long, the carbons burn flat. Hence 


FIG. I. 


certain point, when in the open 
the current drops quite suddenly. 

If the voltage is still increased the 
current will again become steady at a 
much lower value. Between the 
values before and after the drop, the 


Angle down from Horizontal 







r •■». a 







75 .3 









~ r 248 















1 440 























' 1,040" 

» 295 








* 73 . 25 

Green C. P., both globes on 

■ 1 ,060 

Red C. P., both globes on 

*■ 230 

Red C. P. , naked arc 



arc is unstable and hisses. The hissing 
period is indicated by the dotted line 
in Fig. I. 

The hissing point being absent in 
enclosed arcs, the curve for such arcs 
is quite continuous. 


In any circuit an increase of e.m.f. 
must produce an increase of current. 
Hence, as the arc characteristic shows 
a decrease of current with increasing 
e.m.f. there must be placed in circuit 
with the arc a resistance great enough 

FIG. 2. 

to compensate for this tendency if the 
arc is to be kept in a stable condition. 
The method of calculating this re- 
sistance is given below : 

e;=OE=generator terminal volts. 

v=arc terminal volts. 

a=resistance (apparent) of arc. 

x=resistance in circuit, exclusive 
of arc and generator resist- 

i=current, amperes. 

(i) -=a 


(2) — =a+x 


(3) v=ia 

(4) e=ia-j-ix 

(5) e=v+ix 
e — v 

(6 ) =x 

On the arc characteristic (Fig. 2), 
find the point P corresponding to co- 
ordinate v and i, and set off OE=e 
along the vertical axis. Then join 
EP and continue EP until it cuts the 
horizontal axis at x. 

If EX cuts the curve not only at 
P, but at S, above P. it would appear 
that the e.m.f., e could support either 
of two arcs. The point S, although 
affording a mathematical solution, is 
not physically possible, as it would in- 
volve an increase of e.m.f. producing 
a decrease of current, as may be seen 
bv moving EX upward, parallel to its 
original position. Hence the arc cor- 
responding to P is the only one pos- 

January, 1908 



If EX cuts the curve not only at 
P, but at some point, say, T, below 
P, the same reasoning- shows the arc 
at P to be unstable, while that at T is 

Hence for any point P, the arc is 
stable only if the line EX does not 
cut the curve below P, and, therefore, 
if the resistance x be calculated so 
that EX is tangent to the curve, x 
or any greater resistance will steady 
the arc. 


E=generator e.m.f. 

V=potential difference between 

carbons just before current is 

increased to hissing point. 
A=max. amperes, which will not 

produce hissing. 
D=drop in volts from silent to 

R=rise of current from silent to 


Thus, if the generator e.m.f. is 
great, and therefore the steadying re- 
sistance great, the rise of current at 
hissing will be less than when the 
e.m.f. is small. 


The increase of candle-power with 
reduction in size of carbons is well 
illustrated by the following table. It 
will be noted that the candle-power 
can be greatly increased by reducing 
the size of the carbons, this increase 
being obtained without the expendi- 
ture of an additional amount of elec- 
tric energy. The consumption of the 
carbons is more rapid with small car- 

tive carbon is the principal source of 
light in the ordinary carbon arc, the 
bulk of the light from a chemical car- 
bon arc emanates from the flame, and 
is apparently due to minute burning 
particles in the flame, which are raised 
to a very high state of incandescence. 
It has been found that the relative 
brilliancy of the flame of a chemieal 
carbon lamp is about one-third that 
of the positive and negative craters. 
It must be remembered, however, that 
the area of the flame visible at any 
angle is many times that of the crater, 
and the total light emitted by the 
flame is consequently many times that 
emitted by the craters. 

"The carbons are usually of the 
composite type, consisting of three 
zones. The outer zone, or envelope, 
is composed of pure carbon, giving 
mechanical strength. The next con- 
tains carbon mixed with various salts, 
such as those of calcium and magnesi- 
um and the inner soft centering core 
is made of the same materials less 
strongly compressed. 

"The carbons are alongside of one 
another, instead of being coaxial, and 
are inclined, so as to bring their tips 

fig. 3. 


near and pointing downward, the 
craters being so located that the car- 
bons do not obstruct any of the light. 
"Being free from shadow, the 


> 3.5 AMPERES. 


Size of Carbons (Inches) 


















































































[G. N. Eastman (.Elect. World and En?,., April 15, 1905.)] 


bons than with those of normal size, 
so that it is not desirable to reduce 
the size of carbons without 
to the cost of trimming. 


(Abstracted from paper by L. An- 
drea's, Inst. Eire. Eng., Aug., 1906.) 
"Whereas, the crater on the posi- 

opalescent globe of a flame arc looks 
like a globe full of light. There is 
a certain amount of flickering, which 
is, however, not unpleasant for out- 
door illumination. The fumes given 
off by the burning chemicals make 
the lamp unsuitable for use in a room 
not efficiently ventilated. 

"The carbons are at present con- 

siderably more costly than pure car- 
bons, and as the globes cannot be en- 
tirely enclosed, the life of the car- 
bons is very short. Owing to its 
yellow color, the chemical carbon arc 
is useless where discrimination of 
colors is required, but for the illumina- 
tion of the outsides of public build- 
ings, theaters, etc., there is no light 
so pleasing. 

"These arcs work on direct or al- 
ternating-current circuits, their su- 
periority over the ordinary alternating- 
current arc being especially notice- 
able on account of the decreased flick- 
erings at low frequencies, and the gain 
in efficiency due to the two craters 
pointing downward — the positive 
crater being concealed during half of 
every cycle in an ordinary arc. 
Taking the mean of the results 
given by different authorities, the effi- 
ciency of the chemical carbon arc is 
about 5.8 candles per watt." 

If the two carbons are placed, as 
usual, one above the other, the light 
is very unsteady, due to the rising and 
whirling of the vapors ; moreover, 
slags form on the upper carbon and 
drop on the lower carbon, tending to 
extinguish the arc. 

The salts usually employed are cal- 
cium fluoride and other similar fluor- 


(Mr. Gastcr, Inst. Elect. Eng., May. 

The essential feature of the lamp 
is that the lower carbon stands in 
mercury amalgam, and when the lamp 
is switched on a luminous arc is at 
once formed, which heats the lower 
carbon and evaporates the mercury. 
The evaporation requires only a few 
seconds, and increases the luminour 
arc and the intensity of the light, as 
the radiant light of the hot mercury 
vapor is added to the light of the arc. 

As the mercury vapor, togethe - 
with the luminous arc, are enclosed 
in a glass globe, the vapor cannot es- 
cape, and when condensed on the wall" 
of the globe is returned to the genera', 

The lamp will burn for 1200 to 1600 
hours with a single pair of carbon - 
and yields an intense light from 30c 
to 30,000 c. p. at a consumption 
of .2 to .4 watts per candle. The in 
ventor gives for a lamp consuming ir 
amperes at 50 volts and using 14 mm 
carbons, an hourly consumption o 
.25 mm. for the positive electrode am". 
.1 mm. for the negative electrode. 


Watts per mean hemisphcrica' 
c. p. = .299: Amps. =8.7; volts acrcrs 
arc=44; m. h. c. p. =1352; diam. 01' 
carbons, 9 mm. positive and 8 mm. 

Electrician, April 7, 1905. 



January, 1908 


One of electrodes consists of black 
oxide of iron mixed with salts of 
chromium, titanium, etc. The posi- 
tive electrode is copper. 

The color is very close to daylight, 
and carbons run from 1 50 to 200 hours 
for one trimming. Efficiency more 
than .5 watts per candle. 

The positive electrode is made so 
massive that it conducts away most 
of the heat from the arc and does not 
become very hot, thus forming a 
permanent part of the lamp. 

The negative electrode burns at the 
rate of about y$ inch per hour. 

All the light comes from the column 
of vapor, which is from }4 to \ l /% inch 
long. During burning a fine smoke is 
given off, which is conveyed away by 
a chimney ; the lamp is therefore not 
suited at present for indoor work. 


An arc light is now being developed 
on the lines of the magnetite arc, 
titanium oxide being, however, used 
in the negative electrode. This being 
the most luminous substance known, 
promises to give a light of very high 


This is a long-flame arc with in- 
clined carbons, the peculiarity of 
which is the magnetic control of the 
arc. which makes the light far steadier 
than that of the chemical carbon arcs. 
The color of the light very closely 
resembles that of daylight. 


An arc with abnormal actinic power 
for photographic work may. according 
to A. Kutferrath, be prepared by im- 
pregnating the carbons with a mixture 
composed of equal parts of yttrium 
and lead nitrates. 


(P. Cooper Hewitt Lamp.) 
The mercury vapor lamp consists 
of a glass tube exhausted of air and 
containing a small quantity of mer- 
cury. This mercury is connected by 
platinum leading in wires to the nega- 
tive main, and a platinum electrode 
at the opposite end of the tube to the 
positive main. When the arc or 
vapor column is established, the mer- 
cury boils and is recondensed and used 
over and over again. 

The color of the arc is an intense 
vellowish green, which renders the 

lamp useless for general illumination. 
It is, however, finding great favor for 
out-of-door illumination, especially in 
parks and grounds abounding in green 
foliage, and is also largely used for 
factory and workshop illumination and 
for photographic work 


FIG. 4. 

As in the carbon arc light, this is 
accomplished by having two electrodes 
which, when brought together, estab- 
lish a current which supplies the en- 
ergy necessary to build up the vapor 
bridge or arc. The electrodes usually 
consist of mercury, and are brought 

-(^) — aaaaa/ww 


FIG. 5. 


together either by tilting the container 
or moving suitable parts from outside 
by a magnet. 

Starting may also be accomplished 
by applying an excessive voltage be- 
tween terminals. While, however, the 
operating voltage increases directly as 
the length, the starting voltage in- 


B is Lrght Efficiency Curve showing 
relutipn bet .vtvii Camllt i ".v 1 ; Lamp. 
' watts and Current. \/ 

/A is a Direct-Current Lamp Characteristic 

showing relation t»< tween Current and 

Voltage Drop Aeross fypfe KLamp. 






creases more nearly as the third power 
of the length. 

In the alternating-current lamp, 
shown in Fig. 4. the starting is ef- 
fected by means of a small electrode 
or pin placed in the head of the lamp 

and connected to one of the positives 
through a rather high ohmic resist- 
ance. In starting, the tube is tilted so 
that the mercury forms a continuous 
stream from the negative to the posi- 
tive end and is carried by its mo- 
mentum into contact with the pin. 

On account of the irregularity of 
the flow of the mercury it here makes 
and breaks contact with the pin a 
number of times, each time causing 
a breakdown of the negative electrode 
resistance, either on the column of 
mercury or the pin. In the latter case, 
the lamp will go out at the end of the 
alternation. If, however, the me- 
chanical break at the pin occurs dur- 
ing such an alternation that the mer- 
cury column is the negative the lamp 
will start to operate upon the pin, and 
that main positive electrode to which 
the pin is not connected as positive 
electrodes and the mercury stream as 
the negative. Then, on account of the 
starting resistance connected with the 
pin. the current will be immediately 
transferred from it to the correspond- 
ing positive, and the lamp is started. 


As in all other arcs, there is a rapid 
evaporation from the negative elec- 
trode, which gives rise to violent agi- 
tation on the surface of the mercury, 
and at the point of maximum activity 
causes a marked depression. The 
evaporated mercury is cooled and con- 
densed by contact with the bulb, on the 
inside of which it collects in drops. 
The drops grow larger until they run 
down into the electrode. In con- 
verters the main object is to get as 
much cooling surface as possible with 
the shortest practicable vapor path, but 
concentration of heat and a long arc 
arc essential for lighting. 


With small currents there is a ten- 
dency for the arc to fluctuate, and it 
is usual to provide a choke coil to 
keep down the fluctuations. This 
tendency practically disappears with 
currents over 4 or 5 amperes. 

Only choke coils with open mag- 
netic circuits can respond quickly 
enough to be of service in counter- 
acting this impulse. The capacity of 
the wires between choke coil and tube 
should be as slight as possible, as it 
ha:- been found that a twisted pair 
of insulated wires, 10 feet long, has 
a perceptible weakening action on the 
coil if connected between it and the 
negative electrode. 

Systematic Testing of Oil in Transformers and 
Methods of Reclaiming Oil for Service 

H. N. N. 

OWING to the fact that several finger is released for a moment until serial and company numbers of the 
transformer breakdowns were the oil in the sneak is at the same transformer are then noted, together 
found to be caused by the poor level as in the transformer. The sneak with the number of the jar, also condi- 
condition of the transformer oil, it is again closed and rapidly drawn out, tion of case, whether leaky or dirty, 

and condition of terminals, whether 
dirty or loose. This is repeated, using 
a jar for every transformer sampled 
until all jars are filled. 

It is important that both sneak and 
jars be clean before taking any sample 
from a transformer, as dirt or mois- 
ture may be carried over from a previ- 
ous sample. A piece of cloth, which 
will not fur or fray out, is satisfac- 
tory for the purpose ; for cleaning the 
sneak it may be attached to a rod or 
stiff wire and the tube cleaned in a 
similar manner to a gun barrel. 

After the necessary number of 
samples have been obtained they are 
taken to the laboratory and tested. 

The testing arrangement consists 
of a step-up transformer (no to 22,- 
000 volts), and a variable impedance 
connected in series with the low-ten- 
sion windings, which are protected by 
circuit breaker and fuses. The oil- 
testing cup is of the standard design 
made by the Westinghouse Electric & 
Manufacturing Company of 200-c.c. 

FIG. I. 


was decided to begin a regular in- 
spection of the oil to determine its con- 

A sample kit was built containing 
two shelves, each holding six pint jars. 
Each jar is numbered and held by 
spring clips mounted on the back. The 
shelves are provided with separate 
doors which swing through an angle 
of 90 degrees, allowing easy access to 
the interior. The kit is provided with 
a handle and is of such shape that it 
may be easily carried. When filled, it 
weighs 28^4 lb. 

Provided with this kit, a note-book 
and an oil "sneak," which is merely 
a tube long enough to reach to the 
bottom of any transformer, and of 
such diameter that one end may be 
closed with the finger, the oil in- 
spector goes from station to station. 

After removing the cover from the 
transformer to be sampled, and ma- 
king sure that both oil jar and sneak 
are clean, one end of the sneak is closed 
with the finger and the sneak im- 
mersed in the oil, taking care to keep 
clear of the terminal boards, so as to 
avoid a short circuit if the apparatus is 

When the sneak strikes bottom the 


its lower opening brought over the 
mouth of the jar and the finger re- 
leased, thus emptying the sneak. This 
is repeated until the jar is filled. The 

The breakdown voltage, or rather 
the disruptive strength, is measured 
by means of a standard spark gap be- 
tween needle points. The needle 




January, 1908 

spark gap is adjusted to 0.725 in., 
which is equivalent to 15,000 volts 
(sine wave) breakdown. The needle 
spark gap and the oil-testing cup are 
connected in parallel to the high-ten- 
sion terminals of the transformers. 

just below a dull red and then plunged the transformer is thoroughly cleaned ; 

into the oil. Any hissing or crackling 
sound denotes the presence of 

After these tests have been 
pleted the operator fills out an 



all contacts gone over and any neces- 
sary repairs made to both case and 
core, and then refilled with approved 

When using the pump for this pur- 
pose a small amount of oil should be 
drawn off before discharging into the 
transformer, in order to remove any 
dirt which may have been lodged in 
the pump while emptying the trans- 

The drums containing the defective 


Care of- 


When making a test the oil cup and 
electrodes are thoroughly cleaned and 
the cup filled with oil from one of the 
sample jars to the 0.198 c.c. point. 
The upper electrode is then inserted, 
which will cause the oil level to rise 
to 0.200 c.c. The sparking distance 
is adjusted by means of a micrometer 
screw to 0.15 in. 

After the oil has been allowed to 
stand for about five minutes in order 
to free it from air bubbles, the im- 
pedance being all in series with the 
transformers, the current is thrown on 
and the potential gradually increased 
until the oil gap or the air gap breaks 

If the oil gap breaks down the oil 
is classed as defective, because its 
breakdown point is below the limit 
(15,000 volts per 0.15 in. gap). If 
the air gap breaks down it shows that 
the oil will stand 15,000 volts, and is in 
satisfactory condition. 

In case of doubt, the oil cup should 
be emptied and the test repeated with 
a new supply from the same jar. The 
oil cup and electrodes should be thor- 
oughly cleaned and the needles re- 
newed between tests. 

After the breakdown test has been 
completed about a half pint of each 
sample is put into a clean cup and 
tested for moisture. For this purpose 
a one-pound soldering copper will an- 
swer admirably. It should be used for 
no other purpose. It is heated until 

Note. — C. C. stands for cubic centimeters. 

Test Report," to which he transfers all 
data. The reports are sent to the en- 
gineering department and noted. In 
case the oil does not come up to the 
requirements, orders are issued to 
have it replaced. 

The transformer is then cut out 
of service and the oil drawn off into 

Form 280-0-7-OT 






oil are then labeled in a proper man- 
ner and sent to the laboratory. The oil 
is passed through a combined filter and 
dryer and is discharged into clean 
drums where it is allowed to settle, 
and. just prior to shipment, another 
sample is drawn from the bottom of 
each drum and tested as described 
above. If it passes this test the 
sample is preserved and properly la- 
beled until the drum from which it 
was taken has been emptied. 

The filter used by this company is a 
standard oil-cleaning device, and 

Transformer Oil Test Report. 

Tested by ^....M^/UiA<L^... Date ^...T.2. ~ ~ 190 "7 

Kind of Oil &^jX^l<C^Jr. 

From VsU<\XM c^t^/w... Station 

Transformer Co. No &/./.&- 3 Serial No. T«? T V 6 




....... t .j 

Note; Striking Distance on Standard Test to be 0.15 Inch 

steel drums. This may be advantage- 
ously done with a small centrifugal 
pump. As soon as part of the trans- 
former core is exposed the discharge 
may be turned on the core and all de- 
posits of dirt washed off and the oil 
ducts cleaned. 

After the old oil has been removed, 

equipped with resistance coils wound 
on a jacket surrounding the oil reser- 
voirs for the purpose of raising the 
temperature of the oil above the 
evaporation point to obtain moisture 
at the time it is circulating through 
the filtering cells. 

Complete filtering and dehydrating 

January, 1908 



equipment on a self-contained base 
and equipped with motor-driven 
pump is now manufactured by the 
Westinghouse Electric & Manufac- 
turing Company, but at the time this 
work was planned by us and the ap- 
paratus purchased, such a special de- 
vice was not available. 

The drums are cleaned by the fol- 
lowing method : 

The drum is mounted in a frame 
between two pivots in such a manner 

that it may be revolved axially at 
about 30 rev. per rain, by means 
of a small motor. Before start- 
ing the motor several handfuls 
of slugs or nuts and several quarts of 
gasoline are placed in the drum. It 
is then rotated until the scale and 
dirty oil have been removed. 

The estimated cost of inspection, 
test, filtration, and all charges inci- 
dental to handling and transportation, 
is about two cents per gallon, this be- 

ing roughly about 10 per cent, of the 
cost of new oil. 

This system has been in operation 
about one year, and 85 per cent, of all 
transformers tested were condemned 
and the oil changed and filtered; of 
these, none have shown breakdowns 
or discharges through the oil. 

The defective oil drums are labeled 
with a red label, and the filtered oil 
is labeled with a similar white label, 
as shown in the illustration. 

Novel Single Phase Textile Motor Installation 

THE three-phase induction motor the inherent advantages of the three- 
has long held an enviable posi- phase induction motor have long been 
tion in the operation of all known and appreciated, but while 
classes of machinery on account of its their use has been confined to the 


excellent starting characteristics and 
its ability to give constant and reliable 
service with a minimum of attention 
and repairs. The single-phase motor, 
on the other hand, has not been so 
fortunate in gaining the attention of 
both manufacturers and power sta- 
tions on account of its poor starting 
characteristics, and for years the 
manufacturers of single-phase motors 
have centered their efforts on the pro- 
duction of a motor having all the de- 
sirable features of the polyphase 

The General Electric Company has 
perfected a single-phase motor that is 
remarkably free from the defects 
usually found in single-phase motors, 
and which bids fair to become a suc- 
cessful competitor of the polyphase 
motor in the field of small motor- 
driven machinery 

advent of the single-phase induction 
motor has opened up a new field for 
the textile industry: viz., the opera- 
tion of mills in the suburbs or sections 
of the city supplied only with single- 
phase current for lighting purposes. 

An interesting example of the appli- 
cation of single-phase motors to the 
operation of textile machinery, and 
one that is worthy of notice, is found 
in the factory of the Ayvad Mfg. Co., 
of Hoboken, N. J. The product of 
this company is the well-known 
"water-wings." a device used by 
bathers and beginners in swimming to 
assist in floating the body. 

The style of drive used throughout 
is what is commonly known as the 
group drive, in which a number of 
machines are driven by a motor from 
counter-shafts. The electrical instal- 
lation, consisting of nine single-phase 
induction motors, of a total capacity 
of 7SY^ h- P-> was furnished by 
the General Electric Company. The 
type of motor used is one recently 
developed by the above company for 
operation on single-phase circuits and 
is known as the Form KG. 

The conditions influencing the final 
decision in favor of the single-phase 


operation of mills supplied with poly- 
phase current from either central 

alternating-current motors were that 
the factory was located in a part of 

For the driving of textile machinery power station or isolated plants, the the city supplied with single-phase al- 



January, 1908 

ternating current for lighting, and 

that the motors used must operate 
without affecting- the lights on the 
same circuit ; also that all sparking 
of motors must be entirely eliminated. 

A wad Mfg. Co. was to buy the cloth 
used in the manufacture of their prod- 
uct from textile mills, their factory 
converting the fabric into the finished 
product. Since it was impossible to 

supply the demand. Any excess of 
cloth produced forms a profitable side 
line. The factory when in full opera- 
tion will furnish employment for 
about 6o persons, and will require 

FIG. 4. — 10 H.r. 



flaws, there often re- 

This last condition is one peculiar to suited a product of inferior quality, 

textile mills, for in the early stages of Delays in shipment and in transporta- 

cloth manufacture the cotton is in a tion further added to the difficulties of 

highly inflammable condition, and the manufacture. 

least spark might start a disastrous To avoid all the delays and to in- 

fire. A thorough investigation of the sure a good uniform quality of cloth. 


single-phase motor finally selected 
satisfied the manufacturers that all 
conditions were met in the design of 
the motor, all commutators and 
brushes being eliminated. 

Until this year the practice of the 

a complete textile equipment from 
picker to loom was installed. The 
textile equipment now enables them 
to operate the factory during the en- 
tire year, where heretofore only six 
months' operation was sufficient to 


an average of about 5000 lb. of cot- 
ton weekly. 

The view of the picker room, as 
shown in Fig. 1, gives the method of 
counter-shaft connection to the motor. 
The power required for running the 
picker is about 7^ h.p. The picker 
room was built as an annex to the 
main factory, and to provide sufficient 
light for working a skylight was 
placed in the roof and the walls 
painted white. 

The two 10-h.p. motors shown in 
Fig. 2 drive separate line shafting, 
from which are driven the carding ma- 
chines, drawing frames, roving ma- 
chines and fly frames. In this same 
room the lines from the power circuit 
enter through the necessary switches, 
protective devices, measuring instru- 
ments, etc. 

In Fig. 3 is shown a view of the 
spinning room and two of the three 
10-h.p. motors driving 12 spinning 
frames with a total capacity of 3000 
spindles. The third motor, not shown 
in the view, drives a spooler and one 
warp machine. All motors are in- 
stalled on platforms suspended from 
the ceiling and drive the spinning 
frames through countershafting. 
Sheet-iron arches over the passage 
ways protect employees from accident- 
al contact with the belts. In the back- 
ground of the picture may be seen the 
starting-boxes for the motors. All 
the starting-boxes are installed on 
slate panels with a backing of sheet 

In the view of the loom room shown 
in Fig. 4 may be seen the 10-h.p. 
motor driving the looms. It was cal- 
culated that about 7$^ h.p. would be 
required to drive the looms and a 
slasher not shown in the view. This 
leaves an ample margin in additional 
horse power should it be found neces- 
sarv to install other looms. 

January, 1908 



The 5 h.p. shown in Fig 5 drives 
a pony cylinder printing-press, also 
line shafting from which are driven 
cutting presses and other special ma- 
chines used in shaping and finishing 
the "wings." The printing-press is 
used for printing the design on the 
cloth, and also for printing advertis- 
ing pamphlets and circulars. On this 
same floor a 3-h.p. motor drives a 
group of special sewing machines. 

The entire plant is well equipped, 
roomy, and modern in every respect. 
While an excellent example of the 
methods in which single-phase motors 
may be advantageously applied to the 
operation of textile mills, it also 
brings forward to central-station man- 
agers the possibilities existing in ter- 
ritories reached by single-phase dis- 
tribution lines. 

It is expected that the adoption of 
the electric drive at this plant will re- 
sult in a maximum output with mini- 
mum operating and maintenance ex- 
penses, as has been proved in similar 
cases, not to mention the additional 
advantages of cleanliness and flexi- 
bility that go far toward making this 
method of machine drive popular with 

The Rosenberg Generator 

The distinctive characteristic that 
renders the machine especially valu- 
able for certain purposes is its tend- 
ency to deliver a constant current at 
variable speed, and a constant output 
at constant speed. The means taken 
to secure these results are very simple 
and eminently effective ; they consist, 
essentially, in short-circuiting what 
in an ordinary dynamo would be the 
service brushes, and in placing the 
actual service brushes at points on the 
commutator midway between those of 


S 80 



> 60 



40 80 120 1(50 200 240 280 320 
A Til p. 


the first set. The field cores, at least 
in the case of a series machine, are de- 
signed for a much higher degree of 
saturation than is usually the case in 
an ordinary dynamo, and the pole 
pieces are of different shape and of 
greater size. In appearances the ma- 
chine differs but little from a normal 

Tests made at Schenectady on a 
one-kilowatt Rosenberg generator 
showed a voltage of 30 when the 
speed was 1200 rev. per min. ; the 
speed was then increased to 2600 rev. 
per min., and the voltage rose to 30.5, 
and then fell back to 30 ; thus, with an 
increase in speed of more than 100 
per cent., the voltage increased but 
five per cent., while the variation in 
current was slightly smaller. 

At the beginning of the curve the 
characteristic is ascending (Fig. 1). 

At a small value of the current, 
however, the field cores become highly 
saturated, due to their relatively small 
area and the large number of turns 
upon them, while the iron of the arma- 
ture and that of the heavy pole shoes 
is still at a very low density. Any 

* General Electric Review. 


increase in current above this value 
has practically no effect upon the 
strength of the primary field, but pro- 
duces in the large volume of iron in 
the armature core a counter flux 
which is almost proportional to the 
current, and owing to this condition, 
the machine has for the most part a 
drooping characteristic. 

By suitably dimensioning the vari- 
ous parts of the machine, it is possible 
to obtain a short-circuit current which 
will exceed the normal current by any 
required amount, say 25, 50, or even 
100 per cent., while on the other hand, 
the maximum voltage may be made to 
exceed the normal voltage by a corre- 
sponding percentage ; furthermore, 
the machine may be designed to give 
a drop in voltage almost exactly pro- 
portional to the increase of current. 

The current for a given voltage 
may be reduced to any desired value 
by placing shunts of different resist- 
ance across the series field ; in which 
case the field current will no longer 
be equal to the current at the brushes. 
Fig. 2 shows the effects of connecting 
these resistances in parallel with the 
field winding. The decimal given in 
connection with each curve represents 
the proportion of the total current 
that is flowing in the field. 



— — 








40 80 120 160 200 240 280 
4 m p. 

FIG. 2 

The curve marked 0.9 of Fig. 3 
shows the currents corresponding to 
different voltages across the arc of a 
lamp designed for 60 volts and 200 
amperes, the lamp being connected to 
220-volt constant voltage mains. The 
curve shows the performance of the 
lamp without ballast, when connected 
to a Rosenberg series generator ; the 

generator having a low resistance 
shunt connected in parallel with its 
field windings. In a lamp connected 
in series with a ballast consuming 
only 100 per cent, or less, of the 
amount of power expended at the arc, 
the variations of current are greater 
beyond all comparison than those ob- 
tained with the Rosenberg generator. 
With a complete short circuit, the 
machine will require but little more 
power for driving than when open- 

10 20 30 40 50 GO 70 80 

Vol Id 

FIG. 3 

This last named feature of the gen- 
erator, i. e., the inherent impossibility 
of its being subject to heavy overload, 
opens up another extensive field of 
application for the machine in connec- 
tion with gasoline-electric convey- 
ances, both for railway service and for 
street buy lines ; taking as an example 
of the latter, for instance, the Fifth 
Avenue bus, of New York City. 

The Rosenberg generator may be 
further used as a reversible booster 
placed between two sources of vari- 
able voltage, such as a battery of 50 
cells and a no-volt generator; the 
voltage of the latter vary between 90 
and 130 volts. If the voltage of the 
generator and battery are the same, 
a current equal to the short-circuited 
current flows through the booster, 
since the terminals of the machine are 
at the same potential. If, however, 
the voltage of the generator is greater 
than that of the battery, the current 
through the booster will increase 
slightly above the short-circuited cur- 



January, I9O8 

rent, and the booster will generate a 
negative voltage; in other words, the 
machine will run as a motor, consum- 
ing the difference between the volt- 
ages of the generator and battery. 
Should the voltage of the battery be 
greater than that of the generator, a 
current somewhat smaller than the 

short-circuited current will flow, the 
booster will generate a positive volt- 
age, and the potential between the 
two circuits will thus be equalized. 
The machine, when used as an ordi- 
nary booster for charging accumu- 
lators from constant voltage mains, 
will require no regulation throughout 

the whole range of charging, from 
full discharge to complete charge. 
The current decreases with progress- 
ing charging, either very slowly or at 
any predeterminated rate, depending 
upon the design of the machine, while 
the voltage of the booster adjusts the 
line voltage to that of the battery. 

Tungsten. Series Incandescent Lamps at Grosse 

Point, Mich. 

THE city of Grosse Point, Mich., 
has recently installed a series 
tungsten incandescent street 
lighting system where the advantages 
of this kind of lighting are well exem- 
plified. The station equipment con- 
sists of two 8.8 kw. ^ l / 2 ampere con- 



stant current transformers. One of 
these is held in reserve, while the 
other supplies current to JJ 60 c-p. 
General Electric tungsten series in- 
candescent lamps, suspended from 
artistic iron poles. All wiring is laid 
in conduits to the pole, and wires pass 
up the center of the pole to the lamps. 
One of the interesting features of 
this system is the radial reflector with 

which the lamps are equipped. This 
form of reflector, which was recently 
developed by the General Electric 
Company, is so constructed that the 
light is spread and projected very 
evenly over considerable area instead 
of being nearly all concentrated in a 
circle around the lamp. Fig. 2 shows 
the candle-power distribution of a 40 
c.p. series tungsten lamp equipped 
with a radial reflector. Fig. 3 shows 
one of those reflectors on a lamp. It 
may be seen from the diagram of the 
candle-power distribution that at about 
30 degrees below horizontal the ef- 
fective illumination is 50 c-p., making 
the efficiency at this point about one 
watt per candle-power. 

The series sockets with which these 
lamps are provided are so constructed 
that when a lamp is removed from the 
socket two contact plates of large area 
close together before the lamp is quite 
drawn out of the socket, leaving no 
danger of an open circuit at any time. 

Tungsten series lamps are made in 
32, 40 and 60 c-p. sizes, with current 
ratings of 4, 5.5, 6.6 and 7.5 amperes, 
and are exceedingly hardy on account 
of the heavy short filament. They 
will burn for nearly 1000 hours at 
efficiency of from 1% to i l / 2 watts per 

Another installation of series tung- 
sten lamps for street lighting has 
recently been made in Grand Rapids, 
Mich. The lamps are of 60 c-p., and 
were placed on one of the principal 
streets of the city. Judging from the 

complimentary remarks of the press 
and the city council, the test instal- 
lation has proved highly satisfactory. 
The tungsten lamp with its high 
efficiency should greatly increase the 
use of series incandescent lights in 




suburban and residential districts 
where the thick foliage makes it 
necessary to have the units distributed 
at short intervals to produce satis- 
factory illumination. 

Electric Locomotive — Continued 

H. L. ttlRHXR 

Copyright 1907 


NOW that we have the toy motor 
armature rotating continuously 
in a counter-clock direction we 
will focus our attention on the dis- 
turbance going on in the air gap un- 
der the north pole. The main vortex 
— the field magetism — has a clock- 
wise direction. The current in the 
armature wire under the north pole 
is from front to rear and its vortex 
is likewise clock direction. This 
armature wire is moving in toward 
the axis of the field. It is cutting lines 
of force. This cutting, as Faraday 
discovered, sets up an electric tension, 
or voltage, a tendency to drive a cur- 
rent along the cutting wire. Note the 

FIG. 21. 

curvature of the dent the wire makes 
in the field. It corresponds to a 
counter-clock direction of vortex. 
This corresponds to a current from 
rear to front. That is the point I 
want you to note. The line voltage 
drives the current from front to rear. 
The induced pressure is from rear to 
front. The induced voltage opposes 
the line voltage (see Fig. 21). 

Now, if by some means we could 
increase the induced voltage until it 
equaled the line voltage, why no cur- 
rent would flow, of course. Also, if 
we increased the induced voltage still 
further the armature voltage would 
force a current into the line. We can 
accomplish this result by increasing 
the speed of the armature, say by belt- 
ing it to a steam engine. Note again, 
that when the current flows from rear 
to front the whirl around the arma- 
ture wire is counter-clock, consequent- 
ly, external force has to be applied to 
the armature wire to move it in to- 
ward the axis of the field, whose 
direction is clockwise. The steam en- 
gine supplies the necessary force. We 
have, then, an electric machine whose 
direction of field is ciockwise, whose 

direction of armature rotation is 
counter-clockwise. This armature is 
belted to a steam engine whose direc- 
tion of rotation is likewise counter- 
clock. We see that the induced elec- 
tric pressure in the wire under the 
north pole is from rear to front and 
that when current flows in this direc- 
tion the steam engine has to do work 
to force the wire in this counter-clock 
direction. We also see that if we 
force a current through the armature 
wire from front to rear in opposition 
to the induced voltage, that the wire 
tends to move in the same counter- 
clock direction and would keep the 
engine running with the steam cut off. 
Our toy machine, then, revolving 
counter-clock direction in a clock-di- 
rection field, can play the part either 
of the motor or the dynamo. As a 
motor, it takes current from the line 
and transforms it into available power 
at the pulley ; as a dynamo, it absorbs 
the power of the steam engine and 
transforms it into current. It is the 
same field, the same armature, the 
same commutator, the same di- 
rection of armature rotation. In 
one case we supply current and get 
motion, in the other we supply motion 
and get current. The lines are cut in 
the same way in both cases ; the in- 
duced voltage is the same in both in- 
stances. As a dynamo the induced 
voltage drives a current into the line. 
As a motor the induced voltage op- 
poses the current from the line. 

I stated that we could increase the 
induced voltage by increasing the 
armature speed. We know from ex- 
periment that the induced voltage de- 
pends upon the rate of cutting lines of 
force. We can increase the rate by 
increasing the speed, or by increasing 
the number of wires in series on the 
armature, or by increasing the number 
of lines in the field. We can increase 
the number of lines in the field by in- 
creasing the number of ampere turns 
on the field. However, we are not here 
concerned with the various combina- 
tions of speed, armature winding and 
field strengths. That is the work of the 
designing engineer. He must properly 
proportion these. But it is not neces- 
sary to qualify as a designing engineer 
to be able to see that induced voltage 
depends on the rate of cutting lines 
of force, and that an armature wound 
with a few turns of heavy wire can 
deliver a low-voltage, heavy-amperage 

current, and that an armature wound 
with many turns of fine wire can give 
a high-voltage, light-amperage cur- 
rent. In general, the voltage depends 
on the rate of cutting lines of force, 
and the amperage that can be safely 
carried depends on the size of the 
wire. That is as far as we need go 
into the subject of dyamo design. 


We have just been considering the 
forces exerted in the air gap of the 
dynamo and motor, and saw that for 
a given direction of field and arma- 
ture rotation that the direction of the 
armature current decides whether the 
machine is acting as a motor or as a 
dynamo. Keeping this in mind, I will 
ask you to direct your attention to the 

FIG. 22. 

case of two similar machines opera- 
ting in parallel. Let them be street 
railway generators supplying current 
to the same line. We will assume the 
voltage to be 500, the speed 500 r. p. 
m., the direction of armature rotation 
counter-clockwise, the fields separate- 
ly excited, the direction of the field 
magnetism clockwise and the load 
on each machine 500 amperes just be- 
fore the line switches, opened (see 
Fig. 22). The load has disappeared, 
but each armature continues to gen- 
erate 500 volts, and since they are con- 
nected in parallel to the same bus-bars 
each armature tries to drive a current 
through the other, but as their volt- 
ages are equal, why, of course, no 
current flows. Suppose, however, 
that the governor of engine No. 1 
does not hold the speed constant, that 
engine No. 1 speeds up slightly, say, 
two per cent., the voltage of geneiator 
No. 1 rises in the same proportion. 
Generator No. 1 is now giving 510 
volts. No. 2 is giving 500 volts, conse- 
quently, there is 10 volts unbalanced 
pressure between the two armatures, 
and current begins to circulate. The 
510 volts of No. I drive current 
against the 500 volts of No. 2. You 
can readily imagine that the greater 




January, 1908 

the difference in voltage between the 
two machines, the greater the current 
will be that circulates between them. 
In fact, the amperes can be found by 
dividing the effective volts by the 
ohms resistance. Here the effective 
volts are 10 volts. If we assume the 
resistance of the circuit to be o.i ohm. 
why the current will be 10 divided by 
o.i or ioo amperes. But we are not 
here concerned with the calculation of 
currents. What I want you to note is 
that the same current ( ioo amperes) 
is flowing in both armatures and that 
in Xo. I it is flowing in the direction 
of the induced pressure, from rear to 
front, under the north pole, and in 
Xo. 2 it is flowing against the induced 
pressure, from front to rear, under 
the north pole (see Fig. 2$). Ma- 
chine Xo. 2 therefore is motoring. 
The field strength is the same in both 
machines. The armature current is 
the same in both machines : conse- 
quently, we will be justified in as- 
suming the force exerted on an arma- 
ture wire of machine Xo. I. the gen- 
erator, to be the same as the force 
exerted bv a similarlv situated arma- 

ure of the turning force exerted by 
the motor armature current. A prom- 
brake would give us an accurate 
measure. We can readily see that the 
greater the motor current is the great- 
er the torque will be. Likewise, the 
greater the current delivered by ma- 
chine Xo. I. the generator, the greater 
the force that will have to be exerted 
by the steam engine to drive the gen- 
erator armature around. We saw that 
current began to circulate between the 
two machines as soon as the variation 
in speeds disturbed the equilibrium of 
voltage ; consequently, the greater the 
in speed, the greater the difference in 
voltage, consequently, the greater the 
current, therefore the greater the 
torque. A given current in machine 
Xo. 2 will exert a given torque, and 
accordingly life a given load on the 
elevator to which we supposed Xo. 2 
to be coupled. If a bigger load be 
put on the elevator, a bigger current 
will have to flow or the motor will 
stall. What happens, of course, is 
that the motor armature drops in 
speed until the difference in voltage 
between the motor and the generator 

i- small. As a specific case, let the 
line voltage be 500 and the armature 
resistance 0.1 ohm. Let the field be 
separately excited and the armature 
winding and field strength be such 
that one r. p. m. gives one volt. Len- 
der these conditions, if the armature 
revolves 490 r. p. m.. it will give 490 
volts, which is 10 volts less than the 
line pressure. There is therefore 10 
volts unbalanced pressure. One volt 
will drive one ampere through a re- 
sistance of one ohm. one volt will 
drive 10 amperes through 0.1 ohm. 
consequently. 10 volts will drive 100 
amperes through the 0.1 ohm arma- 
ture resistance. The 100 amperes will 
enable the motor armature to give a 
certain torque. Suppose we double 
the load on the motor. Twice the 
torque will be required, consequently, 
twice the current, or 200 amperes 
must flow through the motor arma- 
ture. This means there must be an 
unbalanced pressure of 20 volts. The 
speed drops accordingly to 480 r. p. m. 
We see. then, that doubling the load 
on the motor simply meant a drop in 
speed from 490 to 480. a change of 



FIG. 23. 

ture wire of machine Xo. 2. the motor. 
In machine Xo. 1 the direction of the 
field magnetism is. as stated, clock- 
wise, and the direction of the arma- 
ture current in a wire under the north 
pole is from rear to front, the direc- 
tion of the whirl around the 100 am- 
peres in this wire is therefore counter- 
clockwise. Consequently, an external 
force must be applied to this wire to 
make it move in toward the axis of 
the field. The steam engine supplies 
this force. In machine Xo. 2 the sim- 
ilarly situated armature wire is carry- 
ing 100 amperes from front to rear 
and is accordingly surrounded by a 
whirl whose direction is clockwise, 
consequently, this wire exerts, as 
stated, an equal force in moving in to- 
ward the axis of the field. This turn- 
ing force exerted by the motor arma- 
ture current is called torque. If the 
steam be shut off of engine Xo. 2 the 
motoring machine will keep the engine 
running. If there were some way of 
shifting the belt from the pulley of 
engine Xo. 2 to the pulley of a coun- 
ter-shaft geared to an elevator, we 
could make the motor lift weights and 
could accordingly get a rough meas- 

is sufficient to drive a current through 
the motor armature that will give the 
required torque. The increased load 
on the motor calls for an increase in 
the motor torque. Increased torque 
means increased current in the motor 
armature. Increased motor current 
means increased generator current. 
Increased generator current means in- 
creased drag on the generator arma- 
ture wires. Increased drag on the 
generator armature wires means in- 
creased steam consumption, conse- 
quently, it is the expansion of the 
-team that does the work after all. 
The force exerted in pushing a cur- 
rent-carrying wire across the magnetic 
field of the generator in a direction 
the wire does not want to go, and the 
force exerted by a current-earning 
wire in traveling across the magnetic 
field of a motor in the direction the 
wire wants to go are merely inter- 
mediate steps in the transfer of heat 
into motion. 


W have just seen that the motor 
responds to an increased load by a 
drop in speed. But the drop in speed 


1 -^— - • 

_ liefer 




" ■> 

■j—i — r^ — g| — 



FIG. 24. 

about two per cent. Had the arma- 
ture resistance been half as great, why 
half as much unbalanced voltage 
would have doubled the current, in 
which case the variation in speed 
would have been, approximately, one 
per cent. In fact, this type of motor 
is practically a constant-speed motor, 
from no load to full load, as long as 
the line voltage remains constant. 

If we increase the line voltage the 
motor will speed up until the unbal- 
anced voltage is just sufficient to drive 
the required current through the arm- 
ature. If the line voltage drops the 
speed drops. Assuming the resistance 
to be 0.1 ohm and that one 
revolution per minute gives one 
volt, and that 100 amperes are 
required to give the necessary 
torque, we see that if the line voltage 
is raised to 1010 the armature will 
have to speed up to 1000 r. p. m. to 
keep the unbalanced voltage at 10 
volts. Likewise, if the line voltage 
dropped to no. the armature speed 
would have to drop to 100 r. p. m. to 
keep the unbalanced voltage at 10 
volts. If the line voltage dropped to 
10 volts the motor would stall, for we 

January, 1908 



assumed 100 amperes to be the cur- 
rent required to give the necessary 
torque. If the motor did make one 
revolution per minute with this 
line voltage, it would generate 
one volt. Consequently, the un- 
balanced pressure would be nine 
volts, so only 90 amperes would flow. 
But 90 amperes could not lift the load, 
so the load on the motor would not 
allow the armature to turn. 

We see, then, that the motor arma- 
ture runs at a speed that allows the 
necessary current to flow. For a con- 
stant line voltage the speed is prac- 
tically constant for big variations in 
the motor load. But if the line volt- 
age varies the motor speed varies with 
it. The torque varies with the cur- 
rent, the speed with the voltage. 
These two facts bring us to our next 
point, the work done in an electric 


In the specific case that I just called 
your attention to, I assumed that one 
revolution of the armature per minute 
gave one volt, and we saw that for a 
line voltage of 500 volts the motor 
armature made about 500 r. p. m., 
and for a line voltage of practically 
1000 volts the armature made 1000 
r. p. m. In fact, the speed was a meas- 
ure of applied volts. We can also 
assume that the armature winding and 
field strength were such that one am- 
pere gave a torque of one pound. 
Consequently, 100 amperes would give 
a torque of 100 lbs. The amperes 
then can be taken as a measure of the 
torque. Hence, if we can measure the 
amperes we know the force exerted 
by the motor, and if we can measure 
the volts we know the speed — the 
space through which the force is ex- 
erted per minute. The product of 
force by the space through which the 
force is exerted per minute is the rate 
of work. The product of the volts by 
the amperes, then, is a measure of the 
rate of work. We, of course, can 
measure the work done per minute 
by the motor by the means of a brake. 
The result is expressed in foot-pounds 
per minute. Foot-pounds per minute 
divided by 33,000 gives, as perviously 
stated, the horse-power. There is an 
exact relation therefore of the product 
of current by volts to horse-power. 
The product of one volt by one am- 
pere is called a watt. The size of the 
volt and the ampere are so taken that 
746 watts equal one horse-power. One 
watt therefore is 1 / 7iS OI a horse- 

The voltage applied to incandescent 
lamps is usually 100 volts. This volt- 
age drives a half ampere through each 
lamp. The product of 100 volts by 
one-half ampere means 50 watts, or 
about y is of a horse-power. Fifteen 

lamps, then, would require a little 
more than one horse-power of energy. 
Fifteen hundred such lamps would re- 
quire a little more than 100 h. p., and 
15,000 lamps would require a little 
more than 1000 h. p. Fifteen thou- 
sand lamps at 50 watts each means 
750,000 watts, or 750 kw. So far as 
the steam engine is concerned, the 
work it does is the same whether these 
750 kw. are used to run a motor or to 
furnish light. 

We see, then, that the product of 
the current by the volts is a measure 
of the rate at which the electric cur- 
rent does its work. The current is 
the measure of the force, and the 
volts are a measure of the space 
through which the force is exerted 
per minute. These facts are easily 
verified by measuring the foot-pounds 
of work done per minute by a motor. 
The indicator and the speed-counter 
are a measure of the work done by the 
steam engine. The ammeter and the 
voltmeter are a measure of the work 
done by the generator. The prony 
brake and the speed-counter are a 
measure of the work done by the 
motor. Increased brake friction on 
the motor means increased kilowatts 
output at the generator and increased 
steam consumption on the part of the 
engine. The motor reproduces, minus 
the losses, the work done by the steam 
engine. If there were no losses in 
transmission the kilowatts output of 
the generator would equal the brake 
horse-power output of the engine, and 
likewise would equal the brake horse- 
power output of the motor. 

In a system without losses every 
horse-power applied to the dynamo 
would produce 746 watts, and every 
746 watts put into the motor circuit 
would develop one horse-power. But all 
machines have friction, all wires have 
resistance ; and there are iron losses in 
the armature cores. All these losses 
mean that a certain amount of energy 
is transformed into useless heat. Con- 
sequently, the foot-pounds per minute 
represented by the kilowatt output of 
the dynamo will be less than the foot- 
pounds per minute applied to the 
dynamo shaft, and likewise will be 
greater than the foot-pounds per min- 
ute developed by the motor shaft. 

We saw that heat energy is stated 
in thermal units, and we see that elec- 
tric energy is stated in watts. Thermal 
units and watts are both measured in 
foot-pounds. W T e have, then, a com- 
mon measure for the energy consumed 
by the steam locomotive and for the 
energy consumed by the electric loco- 
motive. The electrical expression — 
watts equals the product of the cur- 
rent by the volts (W=CXE)— is but 
another way of saying that so many 
foot-pounds of work are done per 
minute. If the motor is doing 100 

h. p., we know that the dynamo is 
doing more than 100, and that the 
steam engine is doing still more than 
the dynamo — enough more to take 
care of the friction of the engine itself 
and the losses in the dynamo, line and 
motor. We can measure the current 
and volts, and so can know exactly how 
many foot-pounds are applied to the 
motor, for, as stated, 746 watts equal 
one horse-power, or 33,000 foot-pounds 
per minute. In fact, we can measure 
the foot-pounds per minute delivered 
to the furnace, to the engine, to the 
dynamo, to the line, to the locomo- 
tive and to the train. All these foot- 
pounds means heat units. Watts 
meaning foot-pounds and foot-pounds 
meaning heat units, we have, of 
course, a measure of the heat gener- 
ated in the circuit in overcoming re- 
sistance of the wire. We can measure 
the current in the wire and can meas- 
ure the voltage required to drive this 
current through the resistance of the 
wire. The product of the current by 
the volts gives the watts lost in the 
wire. These watts translated into 
foot-pounds, and these foot-pounds 
translated into heat units, put the line 
loss in terms of coal pile — terms 
which are not difficult to understand. 


We know from experience that cur- 
rent heats the wire that carries the 
current. Silver is the best conductor, 
but its resistance is so little different 
from that of copper that copper is the 
metal that is generally used. I just 
pointed out that the heat generated 
per minute in the wire can be trans- 
lated into foot-pounds per minute, or 
horse-power, or watts. Watts, as 
stated, means the product of the cur- 
rent by the volts ; consequently, we 
can then find the volts required to 
drive the particular current through 
the particular wire under considera- 
tion. I also stated that a pressure of 
one volt would drive a current of one 
ampere through a resistance of one 
ohm. Ampere will, perhaps, mean 
more to you if you will bear in mind 
that current is used for electroplating, 
and that one ampere will deposit a 
certain weight of silver per second. 
Such an arrangement has been used 
as an ammeter. Ohm will mean more 
to you if you will think of it as being 
the resistance of a copper wire 1000 
ft. long and 0.1 in. in diameter. 
The volt, then, is the pressure used in 
driving an ampere of current through 
the 1000 ft. of copper wire 0.1 in. 
in diameter. 

If we had 10,000 ft. of this wire 
we would have 10 ohms resistance, 
and as one volt is used in driving one 
ampere through each 1000 ft., why 
IO volts, of course, will be required to 
drive one ampere through the 10,000 



January, 1908 

feet. The power will be the product 
of one ampere by the 10 volts — 10 
watts. If we had 100,000 ft. of such 
wire we would have 100 ohms re- 
sistance and, consequently, 100 volts 
would be required to drive one ampere 
through it. If the wire were insulated 
and we connected one end to the posi- 
tive brush of a 100-volt dynamo and 
the other end to the negative, one 
ampere would flow through the wire. 
The product of the current by the 
volts would be 100 watts. The wire 
would weigh 3150 lbs. One hun- 
dred watts are converted into heat in 
this wire every second. This is not 
very much, just twice as much as that 
given out each second by one 50-watt 
lamp. Suppose we had used a 1000- 
ft. length. Its resistance would be 
one ohm, consequently, the 100 volts 
would drive 100 amperes through it. 
One hundred amperes by 100 volts 
means 10,000 watts, or 13.4 h. p., or 
7370 foot-pounds per second, or 9.4 
heat units. The weight of copper is 
31.5 lbs. If the current flowed one 
second the temperature would rise 
3. 3 F. Suppose we had used 100 
ft. The resistance would have been 
0.1 ohm and the current 1000 amperes 
until the wire melted. Had 10 ft. 
been used an explosion would have 
followed, for the resistance of 10 ft. is 
0.01 ohm, and the machine would at- 
tempt to drive 10,000 amperes 
through the wire. If the steam engine 
were powerful enough and the dyna- 
mo heavy enough, the energy ex- 
pended in one second would have been 
1,000,000 watts, or 1340 h. p., or 737,- 
000 foot-pounds, or 940 heat units. 
The 10 ft. of wire weighs 0.315 lbs. 
The heat generated would be suffi- 
cient to raise the temperature of 10 
such wires 3000 F. The single wire, 
of course, would vaporize instantly. 
This energy translated into motion 
would raise one of the tunnel loco- 
motives three feet. We see, then, the 
practical necessity of keeping the heat- 
ing within proper bounds. 

Tables have been compiled giving 
the safe carrying capacity of wires, 
and fire insurance companies are very 
strict in their rules on the subject. 
We saw that a copper wire of 0.1 
in. diameter could carry 10 amperes 
with practically no heating. If we 
want to carry 20 amperes we can use 
two such wires in parallel. If we 
want to carry 100 amperes we can use 
a cable made up of 10 strands of such 
wire, or can use a single wire of an 
equivalent cross-section. The proper 
cross-section for the various currents 
has been determined by experience. 

We see from the above the current 
depends upon the resistance of the 
circuit and the voltage applied to the 
circuit. If we keep the resistance con- 
stant and increase the voltage, why 

the current increases. If we keep 
the voltage constant and increase the 
resistance, why the current decreases. 
The expression is current equals volts 

divided by resistance, of C= — (here 

E stands for volts). From the wire 
tables we can pick out the resistance 
of any wire. Consequently, if we 
know the voltage that is applied to the 
particular length of the wire, we can 
find the current which that voltage 
will drive through the wire. 

Now, since C= — , we can multiply 


both sides by R and get C R= 

Canceling out the R's on the right 
hand we get C R=E. The expression 
means that the product of the current 
by the resistance equals the voltage 
required to drive the current through 
the resistance. Consequently, if we 
know the resistance of a wire, and 
know how much current we want to 
send through the wire, we can find the 
voltage required to drive the current 
through the resistance. 

If we divide both sides of our last 


expression by C, we get = — =R. 

C C 
This means that the resistance of a 
wire equals the volts applied to it 
divided by the current flowing 
through it. 

These three expressions are but 
variations of ohms law, which, in its 
simplest form, is that one volt will 
drive one ampere through one ohm. 
If we know any two of the quantities 
we can find the third. It enables us to 
figure the loss in a transmission line. 
We want, for instance, a certain num- 
ber of horse-power delivered at a cer- 
tain distance from the power-station. 
Horse-power means so many watts. 
Watts means so many amperes at so 
many volts. We can find the size and 
resistance of the line wire from the 
wire tables. The line resistance multi- 
plied by the amperes gives us the volts 
absorbed in driving the current 
through the line. The product of 
these volts (lost in the line) by the 
current gives the watts lost in the line. 
These watts are spent in heating the 
line wire. The generator therefore 
must supply these watts in addition 
to those it delivers to the motor. The 
heavier the copper wire used in the 
line, the smaller the percentage of the 
total power wasted in the line. If we 
wanted to transmit power at 500 volts 
with a 10 per cent, loss, we could only 
allow 50 volts drop in the line. If we 
wanted to transmit it at 50,000 volts 
with 10 per cent, loss, we could allow 
5000 volts drop. One thousand kilo- 

watts at 500 volts means 2000 amp., 
while 1000 kw. at 50,000 volts means 
20 amperes. Our line resistance in 

E 50 

the first case would be R= — = = 

C 2000 
0.025 ori m. Our line resistance in the 


second case would be = =250 

ohms. The wire in the first instance 
would have to be about one square 
inch in cross-section, and the power 
could be transmitted about one-quarter 
of a mile with 10 per cent, loss in the 
line. The copper would weigh about 
8000 lbs. In the second case, the same 
weight of wire 0.1 in. in diameter 
could be used and the power trans- 
mitted 25 miles with the same loss. 
In fact, power is transmitted more 
than twice this distance with this volt- 

Ohms law (C= — or C R=E, or 

R= — ) and the expression for watts 

f\V=CE), and the relation of watts 
to horse-power (746 watts =one 
horse-power) and the foot-pounds per 
minute in a horse-power (33,000 foot- 
pounds per minute=one horse- 
power) and the foot-pounds per heat 
unit (778 foot-pounds =one heat unit) 
enable us to figure not only the line 
losses, but the energv consumption of 
the electric locomotive as well, and 
figure it in the same terms that we 
measure the energy consumption of 
the steam locomotive. We start with 
coal in both cases and end with draw- 
bar pull in both. The passage of the 
energy through the electric form in- 
troduces additional losses, but since 
we know how to measure energy in 
the electric form, means have been 
found to minimize the electric losses 
to such an extent that they are but a 
small percentage of the total losses. 


As a result of your scientific reading 
you know that energv can be trans- 
formed but cannot be destroyed. The 
evolution of this idea is the achieve- 
ment of modern science, and is one 
of the greatest generalizations ever 
made. The law is applied to all 
problems involving- the transfer of en- 
ergv' in anv of its forms, and if the 
result cannot balance the energy equa- 
tion, why the result is not valid. 
Knowing then that a certain number 
of heat units should produce a certain 
number of foot-pounds, and knowing' 
how to measure the heat units and 
foot-pounds, we find that Nature 
charges a commission for every trans- 
formation, and we can find out just 
how much she charges. It is the busi- 

January, 1908 



ness of the engineer to minimize these 
losses. Let us glance at the more 
important ones in the case of the loco- 

The heat units in the coal are stored 
energy from the sun, and, as pointed 
out, they represent a certain number 
of foot-pounds of work. By burning 
the coal we have a means of trans- 
ferring some of this stored energy to 
the steam and by allowing the steam 
to expand we can transfer some of the 
steam energy into motion. By the 
means of a prony brake we can meas- 
ure the power developed by the en- 
gine. Knowing the temperature of 
the steam and its pressure, and the 
weight of the water of condensation, 
we can measure the energy supplied 
to the engine. We can measure the 
losses in the steam line due to radia- 
tion, condensation and wire drawing. 
We can measure the heat carried off 
by the flue gases, the losses due to 
boiler and furnace radiation, the losses 
due to imperfect combustion and the 
losses due to wasted fuel. Now, ac- 
cording to the law of the correlation 
and conservation of energy, we know 
that the foot-pounds developed by the 
engine plus the foot-pounds lost in 
the transformation and transmission 
must equal the foot-pounds repre- 
sented by the coal consumed. We have 
just seen that we can measure the 
work done by the motor, and the en- 
ergy consumed by the motor, also 
that we can measure the line losses 
and the output of the generator. The 
output of the steam engine is ex- 
pended in the dynamo. The work 
done by the motor, plus the losses in 
the motor, line and dynamo, equal the 
work done by the steam engine. Con- 
sequently, the work done by the motor 
plus the whole chain of losses equal 
the energy stored in the coal that was 
consumed in doing the work. I stated 
that these various losses can be meas- 
ured. Now, measurement shows that 
the losses due to the electrical trans- 
formation are small compared to the 
losses in the steam end. For instance, 
a iooo-h. p. engine that can transform 
20 per cent, of the energy supplied to 
it into motion represents about the 
best the art can produce ; but an elec- 
tric motor of the same capacity can * 
transform 95 per cent, of the energy 
supplied to it into motion. The bal- 
ancing of the energy equation then 
tells us that the addition of an elec- 
trical transmission system to a steam 
plant means but a slight increase in 
the losses. It also show us that by 
centralizing the steam plants, by using 
improved methods of steam genera- 
tion and by employing electrical trans- 
mission, we can do the work more 
economically than by applying the 
ordinary steam engine directly to the 
work. It is not surprising, therefore, 

that the electrical locomotive is en- 
croaching in the tunnel work and 
terminal work of the steam locomo- 
tive. Now, the essence of the electric 
locomotive is the series motor. But 
before considering the motor, how- 
ever, I want to call your attention to a 
simple transmission system. 

We will assume two similar, but 
widely separated magnetic fields (see 
Fig. 24). We will assume that there 
is a wire in each field parallel to the 
current that produces the field. We 
will assume further that the wires 
are the same length, that they are 
similarly situated and that the front 
end of the first wire is connected with 
the front end of the second wire, and 
that the rear end of the first wire is 
connected to the rear end of the sec- 
ond wire. Now, if we apply a prime 
mover to the first wire and force the 
wire in parallel to the axis of the 
field, we will find that the second wire 
moves in also. Moving in the first 
wire cuts lines of force. The cutting 
sets up a voltage which drives a cur- 
rent through the circuit, say, from 
rear to front, in the first wire. A cer- 
tain force is exerted in moving the 
wire for it is the case of making par- 
allel currents of opposite direction 
approach. In the second wire it is the 
case of parallel currents of the same 
direction. Consequently, as the same 
current is flowing in similar fields, the 
second wire, as it closes in, exerts the 
same force that is exerted in the first 
wire. The second wire will, for in- 
stance, lift a weight equal to the force 
exerted in the first wire. The faster 
we move in the first wire the faster 
the second wire will move in. But the 
second wire will not move quite as 
fast as the first, for while moving in 
it cuts lines of force, and if it cuts at 
the same rate as the first wire it would 
generate an equal voltage and no cur- 
rent would flow. Now, the current 
that actually flows is equal to the dif- 
ference in voltage generated by the 
two wires, divided by the total resist- 
ance of the circuit. The difference be- 
tween the energy spent in the first wire 
and the energy developed by the sec- 
ond wire is the energy lost in the 
transmission. The steam engine forces 
the wire across the dynamo field. The 
line current drags the wire across the 
motor field. So much for the simple 
transmission system. We have al- 
ready considered the motor and the 
dynamo. I will now call your atten- 
tion to the particular type of motor 
that is used on the electric locomotive, 
which, as stated above, is the series 


The striking thing about the series 
motor is that it gives the electric loco- 
motive the characteristics of the steam 

locomotive. I pointed out that a steam 
locomotive exerts a light pull at high 
speed on a level track, a heavy pull at 
moderate speed on an up-grade and 
powerful pull when starting a train. 
Now, the series motor behaves in the 
same way. Its characteristics result 
naturally from the relation of torque 
to current and the relation of speed to 
voltage. When we were discussing 
torque and speed we assumed the mo- 
tor to have a separately excited field 
of constant strength. We assumed 
the field current to be drawn from a 
battery. Now, in a series motor the 
field winding and the armature wind- 
ing are, as the name indicates, in se- 
ries (see Fig. 25). Consequently, the 
current that produces the field also 
makes the armature rotate. We see, 
then, that any variation in the arma- 
ture current produces a variation in 
the field strength. I pointed out when 
discussing torque that doubling the 
armature current in a constant field 
doubles the troque, also that with a 
constant-armature current, doubling 
the field strength also doubles the 
torque. Now, if the armature current 
and the field strength are doubled at 
the same time, why we get, of course, 
four times the torque. Well, some- 
thing like that takes place in a series 
motor. The motor could be so de- 
signed that doubling the current 
would give four times the torque, and 
trebling the current would give nine 
times the torque ; but it is not so de- 
signed, for such a design would re- 
quire an excessive amount of iron in 
the magnetic circuit. The relation is 
true for the light currents, but the 
ratio magnetism to current falls off as 
the current gets heavier. 

We will, for simplicity, neglect the 
tendency of the iron to approach sat- 
uration as the magnetizing current in- 
creases, and will assume that doubling 
the armature current doubles the num- 
ber of lines in the field. We will as- 
sume also that the motor is running 
on a 500-volt circuit and that before 
the current changed the speed was 
about 500 r. p. m. The armature was 
probablv generating about 490 volts. 
Now, if the armature speed did not 
change when the field strength 
doubled, why the armature would gen- 
erate about twice its former voltage. 
But the speed drops, of course. It 
drops to a point where the armature 
generates less than 500 volts. Conse- 
quently, it will drop to at least one- 
half its former speed. But in order 
that double the current may flow the 
armature must generate less voltage 
than formerly, consequently, the new 
speed will be a little less than half 
the former. Likewise, if three times 
the current gives three times the field 
strength, why the speed would drop 
to a little less than one-third the form- 



January, 1908 

er speed for this current. Also, had 
the current been cut down to one-half 
the speed would have more than 
doubled. In general, the character- 
istics of the motor are such that with 
a light current the armature speed is 
high and the torque small, and with a 
heavy current the speed is low and the 
torque great. We see, then, that the 
series motor will run fast on the level 
track, where a small torque is re- 
quired, and will run slow up a grade, 
where a heavy torque is required. 

Assuming that doubling the current 
gives four times the torque and halves 
the speed, it follows that when the 
motor encounters a grade that calls 
for four times the torque, it only takes 
double the current from the line. 
Now a constant-speed motor would 
take four times the current to give 
four times the torque, consequently, 
would take four times the power, or 
twice as much as thi series motor 
takes on this grade. The series motor 
adopts its speed to the grade. The 
constant-field, or shunt motor, as it is 
called, tries to run up all grades at the 
same speed that it runs at on the 
level track, which means excessive 
currents on heavy grades. Under the 
conditions we have assumed, if nine 
times the torque were required, the 
series motor would only take three 
times the current, while the shunt 
motor would take nine. We see, then, 
why the series motor is adapted to 
grade conditions. 

When discussing the steam locomo- 
tive I pointed out that the greater the 
draw-bar pull exerted in starting a 
train, why the quicker the train gets 
up to speed. Suppose that nine times 
the normal draw-bar pull is required 
to bring the train up to speed in one 
minute. Assuming that three times 
the normal current will give the series 
motor nine times the normal torque, it 
follows, in this case, that we can start 
the train with three times the normal 
current. Had a shunt motor been 
used, why nine times the current 
would have been required. However, 
as stated above, the series motor is 
not designed to give nine times the 
torque with three times the normal 
current. The three-fold current can 
give, nevertheless, from six to seven 
times the normal torque. 

Now, while it is true that the series 
motor is by far more economical of 
power in starting than the shunt mo- 
tor is, it is also true that even the se- 
ries motor wastes some power when 
starting on a direct-current circuit. I 
pointed out that when the power is 
thrown on the only thing that stops 
the first rush of current is the ohmic 
resistance of the circuit. The stand- 
ard railway voltage is 500 volts. As- 
sume the ohmic resistance of a 125- 
h. p. motor to be 0.1 ohm. Now, if 

the motor without any additional re- 
sistance were thrown on a 500-volt 
circuit, why 5000 amperes would start 
to flow. Assuming that the motor cur- 
rent must not exceed 500 amperes, we 
see that there must be one-ohm total 
resistance in the circuit at the start. 
Consequently, 0.9-ohm external re- 
sistance must be inserted. This ex- 
ternal resistance is called the starting 
rheostat. Until the armature begins 
to rotate all the energy is transformed 
into heat, and 90 per cent, of this 
waste takes place in the rheostat. But 
as soon as the current begins to flow 
through the armature, the armature 
begins to revolve and generate a volt- 
age. As it gathers speed its voltage 
increases. By the time it is generating 
100 volts, the effective pressure is 500 
minus 100, or 400 volts ; consequently, 
if we want to keep the current at 500 
ampers we must cut down the total 
resistance to 0.8 ohm. We do this 
by eliminating a part of the external 
resistance. By the time the speed has 
reached the point that the armature 
generates 200 volts, the unbalanced 
pressure has been reduced to 300 volts. 
Consequently, to keep the current at 
500 amperes the external resistance 
must be cut down to 0.5 ohm. The 
speed keeps on increasing, and by 
the time it has reached the point where 
the armature generates 400 volts, the 
unbalanced pressure is only 100 volts. 
Consequently, the external resistance 
must be cut down to 0.1 ohm. When 
the further increase in speed brings 
the voltage up to 450 volts, why all 
the external resistance will have to be 
cut out to allow 500 amperes to flow. 
If the speed rises to the point that the 
armature generates 475 volts, the un- 
balanced pressure is 25 volts. This 
will only drive 250 amperes through 
the motor resistance of 0.1 ohm. Con- 
sequently, the torque will, according 
to our assumption, only be one-fourth 
as great, and the speed will now in- 
crease at one-fourth its former rate. 
By the time the armature voltage has 
reached 490 volts, the unbalanced 
pressure is only 10 volts. This will 
drive 100 amperes through the arma- 
ture. We will assume that the torque 
has now fallen to a point where it is 
only able to overcome rolling friction 
of the train, consequentlv, no further 
increase in speed will take place. A 
balance has now been attained, so the 
motor runs on at a constant speed. 

During all the time external re- 
sistance was in the motor circuit, en- 
ergy was being wasted in this external 
resistance in the form of heat. In 
practice, these losses are reduced bv 
using motors in pairs, starting with 
the two motors in series and throw- 
ing them in parallel when half the 
line voltage is attained. Nevertheless, 

there is considerable waste even with 
this series parallel arrangement. 

We see from the foregoing why an 
electric locomotive equipped with se- 
ries motor has the same speed and 
draw-bar characteristics as the steam 
locomotive, and why the electric loco- 
motive also has its lasses. However, 
the electric locomotive as a machine 
for transforming electric energy into 
motion is 15 times more efficient than 
the ordinary steam locomotive is when 
considered as a machine for trans- 
forming heat into motion. As a ma- 
chine, the electric locomotive is sim- 
pler than the steam locomotive. The 
main problem is to get the power to 
it. A practical solution of this problem 
where a distance of a few miles is in- 
volved has been found for direct-cur- 
rent work. Long-distance work in- 
volves alternating current. However, 
before considering the alternating- 
current motor I will direct your atten- 
tion to the relative efficiency of the 
steam and electric locomotive. 


I have pointed out that the steam 
locomotive is neither an efficient steam 
producer nor an efficient steam user, 
but that the modern steam plant is 
both. The modern steam plant can 
produce a horse-power hour with half 
the coal that the average steam loco- 
motive takes to produce a horse-power 
hour. The addition of an electric 
traction system to such a plant need 
not add more than four per cent, to 
the total losses. Consequently, the 
electric locomotive fed from such a 
station practically applies the station 
efficiency to locomotive work. This 
means that the electric locomotive can 
do about twice the work of the steam 
locomotive with the same weight of 

Take an example. A good steam 
locomotive burns on an average five 
pounds of coal per horse-power hour. 
Good coal contains at least 12,000 heat 
units per pound. The five pounds 
then represent 60,000 heat units. But 
the conditions under which steam is 
generated on the locomotives are so 
unfavorable that only about 53 per 
cent, of the heat units get as far along 
as the locomotive cylinders. Now the 
locomotive must be able to run at 
various speeds, and must be able to 
give various draw-bar pulls. Conse- 
quently, a high steam-engine efficiency 
under such variable conditions is out 
of the question. In fact, the locomo- 
tive does not deliver to the train more 
than eight per cent, of the energy that 
finds its way into the cylinders. We 
started with 60.000 heat units in the 
coal. Fifty-three per cent, of these, 
or 31,200, get as far as the cylinders. 
Eight per cent, of these, or about 2500 

January, J 908 



heat units (approximately one horse- 
power hour), are actually transformed 
into train motion. This means that 
67,500 heat units, or more than 95 per 
cent, of the original 60,000, have been 
lost in the transformation. In other 
words, a little more than four per cent, 
of the energy stored in the coal is 
transformed into useful work by the 
steam locomotive. 

Now assume that we burn 2.5 lbs. 
of the same coal in a modern power- 
station. The 2.5 lbs. of coal contain 
30,000 heat units. In the modern 
power-station the design of the fur- 
nace and boilers, and the lay-out of the 
feed-water system and steam lines are 
such that 72 per cent, of the heat in 
the coal can actually reach the engine. 
The design of the engine is such that 
17 per cent, of the energy delivered 
to it can be transferred to the dynamo. 
The losses in the dynamo, in the trans- 
mission system and in the locomotive 

can be held below 30 per cent, of the 
total power supplied to the dynamo. 
We started with 30,000 heat units. 
Seventy-two per cent, of these, or 
21,600 heat units, reached the steam 
engine. Seventeen per cent, of these, 
or 3,670 units, reached the dynamo. 
Seventy per cent, of these, or 2,570 
heat units (approximately one horse- 
power hour), appear as train motion. 
This means that 27,400 heat units, 
or approximately 91 per cent, of the 
30,000 with which we started, are lost 
in the transformation. The output of 
the electric locomotive, then, is a little 
less than nine per cent, of the energy 
stored in the coal. The output of the 
steam locomotive we saw to be a shade 
over four per cent, of the energy 
stored in the coal. Both of these effi- 
ciencies are bad enough, but you will 
note that the steam locomotive takes 
twice as much coal per horse-power 
hour as the electric locomotive. You 

will note, further, that the electric 
losses were 1100 out of a total of 
27,400. In other words, 96 per cent, 
of the losses are chargeable to the 
steam and four per cent, to the electric 

The total fuel bill for all the roads 
in the United States is, of course, 
enormous. According to the records 
it was more than $150,000,000 in 
1905. In fact, it is between 10 and 11 
per cent. of the total operating expense 
of the roads. Now, if the fuel bill for 
an important service can be cut in 
two by the substitution of the electric 
locomotive for the steam locomotive, 
why the electric locomotive will event- 
ually receive the serious attention of 
the railway managers. The fuel bill, 
of course, is not the only considera- 
tion, but it is an important one. It is 
a measurable quantity. It is the one 
with which we are directly concerned. 
(To be continued.) 

Electric Motor Connections 

The single-phase induction motor, 
manufactured by the General Elec- 
tric Company, is wound as a three- 
phase motor, with a squirrel-cage 
rotor. It is obviously impossible to 
start such a motor directly from a 
single-phase line. If it is once started 
in either direction by any means what- 
ever, and brought up to nearly syn- 
chronous speed, the motor will con- 
tinue to run in that direction if con- 
nected to a single-phase line. 

In order to bring the motor up to 
speed, the General Electric Company 






T- J 













the phases are not regular, there is a 
definite phase rotation. 

As the machine reaches synchron- 
ous speed, the switch is thrown to the 
up position, which puts the motor 
directly on the line, at the same time 
cutting out the resistance and re- 

In many cases, both resistance and 
reactance coils in the starting box are 
divided into several parts, so that if 
the machine fails to start readily with 
all the resistance and reactance in, it 
is quite easy to change the connection, 



makes use of a device founded on the 
principle of splitting the phase by 
means of resistance and reactance in 
the line. The starter consists of a 
double-throw switch which inserts a 
resistance and reactance in circuit. 
This changes the single-phase current 
roughly into three-phase, which is ap- 
plied at the three terminals of the mo- 
tor, as, for example, at A, B and C in 
the accompanying diagram. While all 



thus cutting resistance and reactance 
out of the line. 

In order to enable these motors to 
start heavy loads, a slip pulley has 
been devised. This pulley runs loosely 
on the motor shaft, allowing the shaft 
to turn freely in the pulley while the 
motor is being started. As the motor 
comes up to speed a centrifugal clutch, 
which is keyed to the shaft, grips the 
inside of the pulley, making the mo- 
tor gradually pick up the load. 


Copyright, 1907 

The History of Electric "War in 


Formely Commercial Manager of the Toledo Gas, 
Electric and Heating Company 

I then classed my solicitors into 
groups, each group to solicit the class 
of business best suited to him. The 
newest men were put on residence 
lighting, as I was working the same 
plan on residences. After he had made 
good at that work, I put him on apart- 
ment houses, then on small business 
houses, and, finally, on the larger ones. 
I reserved the big stores, hotels and 
office buildings for myself. I finally 
had sixteen men of all grades and 
ability. I appointed one man who had 
proved his worth to take charge of 
the residence and apartment house 
solicitors, and another to take charge 
of all business-house solicitors. I had 
one power solicitor who worked di- 
rectly under my supervision. 

Every user's card had a serial num- 
ber. YVe also had a corresponding 
serial number in a book. When I 
gave a solicitor a bunch of cards he 
was charged with those cards. When 
he returned one, the name was erased 
from the book. If he had not secured 
the contract, and couldn't get it, the 
card was given to another. In that 
way we never let go until we had the 

My plan was to secure every con- 
tract possible before we started to 
run our feeders into the business dis- 
trict, because the Toledo Railways 
and Light Company could not do a 
thing to stop us. Sometimes we 
closed the contract at the first visit, 
but generally the user would put us 
off and call on the Railways and Light 
Company and ask for a lower rate. 
Of course, they couldn't begin to cut 
rates one year and a half before they 
thought we would be able to give 
service, so they would naturally re- 
fuse and then it was easier for us to 
get his contract. 

In the meantime, the engineering 
department was rushing their work 
in an almost phenomenal manner. The 
day the money was paid for the Toledo 
Heating and Lighting Company, the 
consulting engineer commenced work. 
The architect had the plans for the 
new power-house all ready for the 
builder, who was ready to start to 

The three-conductor feeders were 
ordered and a contract was let for 

two iooo-kw. three-phase Westing- 
house turbines, as it would require 
four or five months to deliver one. As 
it was necessary to have something 
to carry our winter load, our plan 
was to get our feeders into the busi- 
ness district and give service by De- 
cember ist (it was then about Sep- 
tember i st), the Westinghouse Com- 
pany loaned us a two-phase 180-kw. 
turbo-generator, on which we got an 
immediate delivery. 

The construction of our feeders 
progressed more rapidly than our 
competitors figured on. The Toledo 
Home Telephone Company was prac- 
tically controlled by the same men who 
controlled the Toledo Gas, Electric 
and Heating Company, and it was a 
very easy matter to lease a few empty 
ducts from them ; consequently, there 
was very little to do except run a pole 
line in a few places. About November 
20th we were giving service in the 
heart of the business district. We 
had over one thousand contracts writ- 
ten and were in a fair way to get our 
share of the business. We had to be 
very careful and not overload our 
180-kw. turbo-generator, which we 
connected in a very unique manner so 
as to get the full capacity on our three- 
wire circuit. Of course we could only 
furnish single-phase service with that 
connection, so we could not connect 
any large motors. During December 
we watched the generator very closely 
and every day we would connect up 
a few more small users, though 
we would not take chances on the 
bis: users. At times we ran the ma- 
chine at 150-per cent, load and had no 
trouble whatever. In fact, that little 
generator ran four or five months 24 
hrs. per day and never stopped once. 

Until about November ist every- 
thing went even better than we ever 
hoped for. but then our troubles be- 
gan. Unfortunately, the manager had 
never been in such a turmoil before 
and the strain was too great for him. 
When word was passed to connect 
up every small user as soon as pos- 
sible, we found that we had no meters, 
and none ordered, consequently, con- 
nections were made without meters. 
The manager said we could approxi- 
mate bills. 

The force for making connections 
was so small that they could not keep 
up with the contract department, let 
alone gain on the one thousand con- 
tracts which were alreadv made. The 

connecting force were greatly handi- 
capped because they worked directly 
under the consulting engineer, who 
was in Toledo sometimes once a week 
for a day or two. No one but the 
consulting engineer had authority to 
buy anything, and I have seen the con- 
necting force loaf for several days be- 
cause they were out of wire or fuse 
blocks or some other small neces- 
saries. The connections were so slow 
that I made a strenuous fight to get 
them to employ a superintendent who 
could be in touch with the men every 
hour in the day. This they finally 
did, and went to the extremes by em- 
ploying a man who had no practical 
experience in such matters, who made 
matters worse, if such were possible. 

This condition existed until spring, 
when we installed our first iooo-kw. 
three-phase turbo-generator. Then 
the trouble began in earnest. We had 
run our high-tension 4600-volt feed- 
ers through trees, regardless of con- 
sequences. Every time a mist would 
fall, so would a few feeders. The 
contracting force would leave an in- 
nocent boy to answer the complaints, 
and the rest would "hide out." 

Fortunately, I had not allowed them 
to connect any large users, so we only 
got a large number of complaints from 
small users. Upon investigation we 
found our high-tension lines in a very 
poor condition. They were run 
through the thickest foliage on a tree, 
and the trees in Toledo are very 
numerous. The reader can imagine 
the trouble we had when I have seen 
a feeder burned off by coming in con- 
tact with a twig about the size of a 
lead pencil ; that twig was 12 ft. 
away from the feeder, but the tree was 
blown toward the feeder and the feed- 
er was swinging so that they finally 

After this lesson, the company final- 
ly employed a good superintendent, 
who lost no time in employing a pro- 
fessional tree-trimmer, who had a 
force of men, to trim out the trouble. 
It was a big undertaking, but by cut- 
ting trees and changing the pole lines 
they finally succeeded in reducing the 
"burn-outs" to a very satisfactory 

But our troubles were not over, as 
the linemen went out on a strike be- 
cause the foreman was discharged. 
He was unfamiliar with lighting work, 
having- been on street railwav work, 
and the company wished to replace 


January, 1908 



him with a man who understood light- 
ing. We were so crippled that for a 
long time we didn't have enough men 
to "chase" trouble, and all of our con- 
nections were held up again. 

During all this time, from Decem- 
ber I st to June ist, the contract de- 
partment was "sitting down." I went 
to the manager and told him we would 
not make any campaign after con- 
tracts until they could be connected 
with first-class service, because we 
couldn't get them at a decent rate, 
as the user, when approached, would 
go to our competitor, who would show 
him the names of users who had given 
us a trial, and had quit us on account 
of poor service. This would invari- 
ably tie him up with our competitor 
for a year or more. 

I argued that if the man were not 
approached he would be more likely 
to remain as he was until we could 
approach him with a safe proposition. 
Consequently, I discharged nearly all 
of my men, only keeping two for resi- 
dence work, one for power and one 
for small stores. I was to attend to 
the large users, but I never saw the 
time when I thought it would be safe 
to go after them. I had taken a few 
contracts from personal friends, who 
made me promise not to connect them 
until I was sure of giving them un- 

interrupted service. Our manager 
often insisted that they be connected, 
as he knew the service would be good, 
but I held them off and always con- 
gratulated myself for doing so, as 
something always happened after that. 
At one time the manager gave orders 
to connect a certain hotel against my 
wishes, and I had to go to the pro- 
prietor and tell him not to allow them 
to do it at that time. It is needless 
to say that they didn't connect it, and 
I didn't let them connect it until some 
time in June when it was safe. 

After the company was able to give 
fairly satisfactory service it went 
after business with only one idea in 
mind — "Take it away from the com- 
petitor at any price," and the competi- 
tor seemed to have the same idea, be- 
cause the largest hotel that used about 
40 kw. on the peak made a contract 
with the Toledo Railways and Light 
Company for about $1,800 per year. 
I am very certain that the Toledo 
Railways and Light Company did all 
the wiring (the hotel was being re- 
modeled and there was a great deal 
to do), and furnished all lamps and 

The Pope Motor Car Works made 
a contract with the Toledo Gas, Elec- 
tric and Heat Company for 500-volt 
direct-current at about 1^2 cents per 

kilowatt-hour. The Toledo Gas, Elec- 
tric and Heat Company furnished a 
motor generator, and I have been told 
by one who should know that only the 
output of the motor generator was 
metered. I could have secured this 
contract in September of the year be- 
fore at 23/2 cents for power, but re- 
fused it. 

When the municipal lighting con- 
tract was let in the fall of 1906, the 
Toledo Railways and Light Company, 
to be on the safe side, bid about $45 
per lamp per year all night and every 
night. I do not remember the full par- 
ticulars, but I do know that they got 
the contract. Matters went along in 
that way until last spring, the Toledo 
Railways and Light Company bought 
all the stock of the Toledo Gas, Elec- 
tric and Heat Company, paying for 
same three shares of their stock for 
four shares of the Toledo Gas, Elec- 
tric and Heat Company's stock. The 
Toledo Railways and Light Company 
guaranteed the bonds of the Toledo 
Gas, Electric and Heat Company. The 
Toledo Railways and Light Company 
in order to do this increased their capi- 
tal stock $2,500,000. That was the 
capitalization of the Toledo Gas, Elec- 
tric and Heat Company. It is a com- 
mon report that the promoters got a 
bonus of the other 25 per cent, of 
stock for putting the deal through. 

Questions and Answers 

Any Questions our readers may put to us will be cheerfully 

answered in this column 

Question. — How shall I arrange to 
connect to auxiliary power service in 
case our plant breaks downf 

Answer. — Use a two-pole, double- 
throw switch. The bus bars in this 
case would be connected to the middle 
lugs of the switch and the generator 
connected to the upper lugs with the 
auxiliary service on the lower lugs. 

Question. — Our generator is 
grounded. How can I tell whether it 
is in the field or the armature? 

Answer. — First, disconnect the field 
connections from the terminal block, 
so that it has no electrical connection 
with any other part of the machine. 
Connect one end of the testing appara- 
tus with any part of the iron frame of 
the generator. Touch the other ter- 
minal to either of the free terminals 
of the field windings. If you get a 
"ring," assuming a magneto is being 
used for testing, it indicates a ground 
in the field circuit. Now disconnect 
one by one in turn the different field 
coils until the "ring" is no longer ob- 
tained. Evidently the last one discon- 

nected before the ring stopped is the 
defective spool. Repairs can then be 
made with the spool in place, if an 
inspection shows that the ground is 
visible and can be reached, otherwise 
the coil is to be removed and rewound. 

If, however, the field circuit does 
not give a "ring" then lift all the 
brushes from contact with the arma- 
ture. With one end of the testing cir- 
cuit still grounded on the iron frame, 
now rub the other end along the com- 
mutator. If a ring results the com- 
mutator is grounded. If no ring is 
obtained it shows the armature is free 
from grounds. 

Before connecting up the machine 
again with the outside leads from the 
switchboard disconnected from the ter- 
minal board of the generator take the 
free testing wire and touch the cables 
which connect the brush rigging with 
the terminal board. If a ring is ob- 
tained it will show that the cables are 
grounded to the frame, either at some 
point where the insulation has worn 
off, or at the studs holding up the 
brush rigging. 

Question. — / am the engineeer op- 
erating a steam engine. My employer 
has lately been securing proposals to 
install a direct-current dynamo of 150 
kw. capacity. Our engine indicates 
200 h.p. on the limit of the cut-off, 
which gives lis a small margin for 
overload. Yesterday the salesman of 
another company offered a proposal 
on an alternating-current generator of 
180 kw. He told my employer that 
although he was going to use only 150 
kw. he would have to put in the 180 
kw. to do the same work as a 150 kzv. 
direct-current machine. As I am re- 
sponsible for the steam plant I want to 
knozv why the size is larger, because if 
it takes a 180 kw. alternating-current 
machine to do the equivalent of 150 
kw. our engine is not going to be big 

Answer. — The reason for an alter- 
nating-current generator having a nor- 
mal capacity greater than the direct- 
current is as follows : 

In a direct-current generator the 
voltage rises to its maximum and con- 



January, 1908 

tinues at that value until the switches 
are pulled or the generator shut down. 
The rate of flow of the current or am- 
peres reaches its maximum value in 
the first closing of the switches and 
continues at that value until the end, 
except as at odd intervals the number 
of lights or motors is increased or de- 
creased. The energy given out at any 
one instant consists of a multiplication 
of the voltage by the amperage at any 
instant, giving us the watts energy 
flowing at that instant. In an alter- 
nating-current generator, however, the 
voltage and amperage go from zero 
to maximum and back again to zero 
many times a minute, a 60 cycle ma- 
chine rising and falling 7200 times a 
minute. If the generator is feeding 
incandescent lamps the voltage wave 
reaches its maximum, or peak, at the 
same instant as the ampere wave, and 
we get the momentary number of 
watts by multiplying the two peaks 
together (volts X amperes). 

If, however, any electrical device 
containing a coil is attached to the cir- 
cuit, such as an arc lamp, motor or 
transformer, this device will have the 
effect of so delaying the wave of am- 
peres that when the voltage reaches its 
peak the ampere curve will be only 
one-half or three-quarters complete. 
When the ampere wave reaches its peak 
the voltage curve will have started to 
fall. If, for instance, we had been con- 
sidering a 100 volt, 50 ampere circuit 
the momentary watts at the top of 
the peaks feeding the incandescent 
lamps =100 by 50=5000 watts (5 
kw). When feeding the motor when 
the voltage was at 100 the amperes 
would only have been at the 30 point 
so that the momentary energy would 
be 100 by 30 or only 3000 watts 
(3 kw.). An alternating-current volt- 
meter will always read dead-beat the 
voltage of the peak of the voltage 
curve, and the ammeter the peak of 
the ammeter curve, so that in both of 
the above cases we would always read 
100 volts and 50 amperes. We learn 
however in the first case we had 5000 
watts at one instant, and 3000 watts 
in the second case, voltmeters and am- 
meters registering same in both cases. 
A dynamo supplying current would 
evidently have to supply 5 kw. for the 
lamps and only 3 kw. for the motors. 
Suppose, however, that the actual mo- 
tor load were 5 kw., the same as the 
lighting load. If it were necessary to 
obtain 5 kw. output under the second 
condition and we could only reach a 
three-fifth point on the ampere curve 
at the instant that the voltage was 100, 
evidently we must get an amperage 
curve larger than before so that three- 
fifths of its actual peak would give 
50 amperes. We would have 100 times 
50 or 5000 watts, which is required, 
but if we have boosted our amperage 

wave our ammeter will now register 
the peak of this new wave at 83.3 am- 
peres (50=3/5 of 83). 

At the instant the amperes are ac- 
tually at the 83 peak the voltage will 
be at 60 volts so that at the height of 
the ampere wave we have 60 by 
83.3=5000 watts=5 kw. 

As explained above with the in- 
struments reading only the peak we 
would have an apparent energy of 100 
by 83.3=8.33 kw. We have seen, 
however, that due to power-factor, in- 
stead of over 8 kw. we were actuallv 
delivering 5 kw. Xow an alternating- 
current generator to feed this circuit 
would have an 8.33 kw. rating but the 
engine driving it would only have to 
be a 5 kw. machine (not counting 
losses). If a direct-current generator 
were selected the actual energy de- 
livered would be the same as the ap- 
parent energy and a generator of 5 
kw. would be selected and the same 
5 kw. engine as before would drive it. 

Therefore the representative offer- 
ing the alternating-current propositi* m 
submitted would give figures on a 
8 kw. generator which would deliver 
to your system only 5 kw. In the par- 
ticular case cited, a 180 kw. alternator 
would do 150 kw. work. On this end 
of a power plant the first cost is 
higher. This is offset by the reduced 
cost of operating an alternating-cur- 
rent motor system, which does away 
with commutator troubles and fre- 
quent attention. 

Question. — We have a 100 kw. al- 
ternator installed many years ago with 
a 5 kw. exciter. We have put on 
about 25 kzv. lighting and have been 
running underloaded until lately when 
a large load of motors has been added. 
My wattmeter shows that my output 
is 85 kw. } and yet my exciter is so hot 
I can hardly touch it. When the out- 
put zeas purchased the exciter was 
guaranteed to haze 2? per cent, ozer- 
load capacity. With my alternator 
underloaded why slwuld the exciter 
be so heavily overloaded at this point. 7 

Answer. — If you will read the an- 
swer to the question above you will see 
that the addition of a motor load re- 
quires an increase of the ampere wave 
delivered by the alternator. Quite 
evidently, therefore, you must increase 
the ampere wave of the alternator by 
making more lines of force for the 
armature conductors to cut through 
as they sweep past the field poles. We 
can only obtain this increased satura- 
tion by increasing the number of 
amperes in the field coils. As this 
increased number of amperes must 
come from the exciter it is necessarv 
to have an exciter with a capacity 
sufficiently large for the purpose. As 
your alternator is underloaded and the 
exciter overloaded, it indicates that the 

exciter now installed is too small to 
deliver enough amperes at these low 
power-factors. You should therefore 
get another exciter of about 5 kw. 
which you can parallel, or do away 
with the present one and install one 
of the 10 kw. If you have 60 kw. of 
a motor load with the power-factor as 
low as 65 per cent., not unusual with 
small motors, you have a load of about 
100 k.v.a., or an apparent load of 100 
kw., and current of this value surging 
in your machine without reckoning the 
lighting load of 25 kw. You have 
therefore a field current 25 per cent, 
in excess and your exciter is over- 
loaded by this amount. 

Question. — / am going to zi'ire for 
t:eo-phase motors on the three-wire 
plan. Hoze much greater cross-section 
should the middle neutral have than 
cither of the outside. 

Answer. — About one and one- 
quarter greater in cross-section than 
either of the outside wires, that is, it 
>h<>uld carry 25 per cent, more 

Question. — We have both alterna- 
ting and direct-current in our building. 
The other day I accidentally connected 
a direct-current motor to the alter- 
nating-current circuit. The motor 
rotated, but flashed horribly. Why 
was this? 

Answer. — Any direct-current motor 
can operate 011 an alternating-current 
circuit but only with this severe spark- 
ing which you noticed. In a direct- 
current motor the lines of force travel 
in one direction. The coils on the 
armature just under the collecting 
brushes are in a position where they 
are momentarily traveling parallel to 
these lines of force and therefore not 
cutting them. Without a cutting 
action no current is being generated in 
the coils, while the collecting brushes 
short-circuit them at the commutator. 
Hence this short circuit does no harm. 
If an alternating current be applied to 
the fields so that the coils under the 
brushes lie in a field which is alter- 
nating back and forth rapidly, current 
will be generated in these coils which 
will manifest itself in the form of 
sparking at the commutator. This 
current will be of fractional voltage 
but high amperage. In the Westing- 
house single-phase type railway motor 
resistance leads are inserted between 
the ends of the armature wires and the 
commutators bars, so as to choke down 
this local current and neutralize its 

You should never connect a direct- 
current motor to an alternating-cur- 
rent circuit because it will burn up 
before you can throw off the switch if 
you have much of a load at the pulley 

Review of the Technical Press 

Energ'y in "Wireless Teleg'rapHy 

(Physical Review, September, 1907, p. 200.) 

As a result of an investigation of 
the relative efficiency of several dif- 
ferent types of receiving systems when 
used under various conditions, the 
author advances reasons for believing 
that the energy in wireless telegraphy 
is propagated through the surface of 
the earth and not by means of a free 
ether wave. 

Solenoid in Series witH Resistance 


(Electrical World, January 18, 1908, p. 140.) 

A theoretical article in which the 
author discusses the design of a 
magnet intended to operate in series 
with a resistance comparable to its 
own value. 

The Self and Mutual Inductances 
of Similar Conductors 


(Bulletin of the Bureau of Standards, January, 1908, 
p. 301.) 

A discussion of current formulae in 
in use with the derivation of a num- 
ber of new expressions of value in 
technical work. 

Plain TalKs on Illuminating 

The Illuminating Engineer, January, 1908, p. 795.) 

The cost of maintenance for iooo 
spherical candle-power hours is given 
as follows : 

Arc lamps (enclosed) ... ^ to ic. 
Carbon filament electric 

lamps 3 to 6c. 

Metallic filament electric 

lamps 4 to 20c. 

Nernst lamps Y* to ic. 

Incandescent gas lamps. . . 2 to 5c. 

Reinforced Concrete Chimneys 

(Cement Age, January, 1908, p. 62.) 

According to the author about 400 
chimneys of reinforced concrete have 
been built since 1898, and the large 
majority of them have given satis- 
faction to their owners. 

Present Status of Incandescent 


(Proceedings of the Canadian Electric Association, 

The most valuable part of the paper 
is the tabular statement of the relative 
cost of operating different series street 
lighting systems. 

Electric Power in Coal Mining' 

(Cassier's Magazine, January, 1908, p. 356.) 

A comprehensive and well-illus- 
trated article dealing with pumping, 
winding and coal cutting. 

Three-Phase Sing'le-PKase Trans- 

(General Electric Review, January, 1908. p. 83.) 

It is impossible to take single-phase 
power from a quarter-phase or a 

three-phase system. In the former 
the power delivered changes from 
maximum to zero and back to maxi- 
mum every half cycle ; whereas, in the 
latter the rate of power delivered is 
constant. Therefore, any system, if it 
be capable of transforming from bal- 
anced polyphase current to single- 
phase current, must be capable of 
storing up energy during the interval 
of time when the power delivered to 
the single-phase is less than the power 
received from the three-phase side. 
The static transformer is incapable of 
storing any energy and cannot be used 
for this purpose. 

E.lectric Welding 

C. B. AUEL. 
(The Electric Journal, January, 1908, p. 18.) 

The author describes a novel use 
of electric welding for the repairing 
of defective castings by melting iron 
in a carbon arc made at the point 
which is to be repaired, the arc being 
made from a carbon stick to the cast- 
ing which melts at the point of appli- 
cation of the carbon. The table gives 
an idea of the data for an average 

welding : 


Volts Across 

Line Volts 




, Includ- 


ing Carbon 

126 (open circuit) . 





























Time of weld = 56 seconds. Hole filled = 1J^ in. 
diameter by 2 in. — approximately. Size of carbon — 
\ l A in. by 6 in. 

"Westing'House Reorganization 

A plan for the reorganization of the 
Westinghouse Electric and Manufac- 
turing Company has been practically 
perfected by the reorganization com- 
mittee. The plan in its general out- 
lines has been adopted by the unani- 
mous vote of the members of the com- 
mittee, but is still subject, before its 
adoption, to the consent of notehold- 
ers and bondholders and to the rais- 
ing of $7,000,000 from George West- 
inghouse and his associates by the 
sale of stock to them. 

The principal provision of the plan 
contemplates the creation of a first 
mortgage bond issue of $45,000,000, 
to bear interest at 5 per cent. Of 
the entire issue $18,500,000 will be 
convertible bonds and be offered in 

exchange for the $18,500,000 out- 
standing convertible 5 per cent gold 
bonds. The remainder will be with- 
out the conversion privilege and will 
be utilized to pay off the floating in- 
debtedness and outstanding short-term 
notes and debenture certificates. 

According to the report of Haskins 
& Sells, certified accountants, the 
company had on the date of the re- 
port a floating indebtedness of $14,- 
000,000. It is proposed to fund most 
of this indebtedness by offering to 
creditors the new bonds, dollar for 
dollar, for their claims. Similarly, 
holders of the $1,969,000 debenture 
certificates, of the $6,000,000 6 per 
cent, collateral notes due August 1, 
1910, and the 5 per cent. French loan 
due October 1, 1917, are to be asked 
to exchange these securities, dollar for 

dollar, for the new bonds. The float- 
ing and funded indebtedness of the 
company amounts to between $43,- 
000,000 and $44,000,000. 

In order to provide working capital, 
the plan contemplates the sale of $7,- 
000,000 stock at par. It is understood 
that George Westinghouse and his 
associates have agreed to take this 
stock, or at least such portion of it 
as is not desired by stockholders. Ac- 
cording to the terms of the plan it is 
understood that the raising of the $7,- 
000,000 is to be effected without any- 
thing resembling an assessment on the 
stock. There is now outstanding 
$4,000,000 preferred stock and $24,- 
000,000 assenting stock. The new 
stock will be of the latter class. 

The committee that has been in 
charge of the plan consists of Richard 




January, 1908 

Delafield, president of the National 
Park Bank ; James N. Jarvie, chair- 
man ; Albert H. Wiggin, vice-presi- 
dent of the Chase National Bank ; 
Paul M. Warburg, of Kuhn, Loeb & 
Company ; F. H. Skelding, president 
of the First National Bank of Pitts- 
burg ; Charles A. Moore, of Manning, 
Maxwell & Moore; Neal Pantoul, of 
F. S. Moseley & Company, of Bos- 
ton, and A. G. Becker, of Chicago. 

A. New Recording Millivolt Meter 
and SH\int Ammeter 

Electrical engineers have long felt 
the need for an accurate and sensi- 
tive recording millivolt meter which 
is adapted to practical every-day serv- 
ice as well as for laboratory tests. 
There has also been a demand for a 
recording ammeter of the shunt type 
which can be connected by leads to 
the main bus bar. The shunt system 


is especially economical where heavy 
currents are to be indicated or re- 
corded, as the instruments may be lo- 
cated at a considerable distance from 
the main current, thus saving great 
expense in carrying the main con- 
ductors to the point where the instru- 
ment is located. 

The recorders illustrated herewith 
have been designed to meet these par- 
ticular demands. The two most im- 
portant fundamental features of these 
recorders are a sensitive electrical 
movement of special design, made by 
the Weston Electrical Instrument 
Company, and a new recording system 
using a patented smoked chart so ar- 
ranged that there is absolutely no 
friction between the recording arm 
and the chart. 

These instruments are so sensitive 
that the recording arm will move over 
the whole scale for five millivolts or 
less, making it possible to accurately 
record one ten-thousandth of one volt. 
The graduations on the chart are even- 
ly proportioned over the entire range, 

the same as the Weston ammeter, so 
that even though there is only a small 
current flowing, the readings may be 
as readily taken as if the current was 
the maximum that the instrument 
would record. This feature will be 
greatly appreciated, as there are many 
places where it is desired to install 
instruments for increasing future de- 
mands, and it is important that the 
records be perfectly clear, even though 
the loads are very light when the out- 
fit is first installed. 

The records are made on a novel, 
semitransparent smoked chart, which 
is periodically brought into momentary 

made to operate twice every second. 
When the record is completed the 
chart is dipped in a simple fixative so- 
lution, which makes the record perma- 
nent for filing. 

The recording millivolt meter is 
shown in Fig. i, and Fig. 2 is a re- 
duced photographic facsimile of a 
chart taken from one of these instru- 
ments in connection with electrolysis 
surveys of underground structures 
which are being conducted by the 
Electrical Testing Laboratories of 
New York City. The graduations of 
this chart are arbitrary. It was re- 
volved once in twenty-four hours and 


contact with the end of the recording 
arm by means of a special vibrating 
device. In this way a series of white 
dots are made on the smoked surface 
and these form a continuous line and 
a record is thus made without caus- 
ing any friction between the moving 
arm and the chart. The rate of vi- 
bration of the chart is timed to suit 



the frequency and range of the varia- 
tion in the current to be recorded. 
The usual period of vibration of the 
chart is once in ten seconds, but to 
obtain continuous lines where the 
fluctuations of the current are quite 
rapid, the vibrating attachment is 

was vibrated once every ten seconds. 
The zero position of the recording 
arm was the middle of the scale, so 
that the record might be independent 
of the direction of the current, as, in 
many cases, the direction of the cur- 
rent changes from negative to posi- 
tive during the day. 

It is expected that by using a num- 
ber of these instruments, operating 
simultaneously at different points, 
stray currents in water and gas mains, 
or in any underground structure, may 
be recorded, making it possible to dis- 
cover the causes of trouble and how 
they may be eliminated. 

The recording ammeter is shown in 
Fig. 3 connected to a standard Weston 
ten-thousand ampere shunt, to which 
is also connected a Weston indicating 
station ammeter. This illustration 
shows that the recorder may be readi- 
ly applied to any standard shunt which 
is already in service without disturb- 
ing the indicating instrument at the 
switchboard. As illustrated here, 
leads of almost any desired length may 
be used to connect the indicating and 
recording instruments to the shunt on 
the main bus bar. It is even possible 
to have the recording ammeter located 
in the superintendent's office at a 
great distance from the shunt and the 
indicating instrument located on the 
switchboard convenient for the ob- 
servation of the operator. Such com- 

January, 1908 



bination outfits could be furnished as 
units, with leads of the proper lengths 
to suit the individual cases. 

The recording shunt ammeter has 
been successfully applied for taking 
continuous records of the current on 
a large trolley system, where the 
fluctuations are very rapid, and varied 
as much as four thousand amperes 
several times in a minute. The charts 
for such work as this are made to re- 
volve once in one hour and the vi- 
brator operates twice in one second. 
For preliminary tests, the recorders 
are provided with special fast vibra- 
tors for the smoked chart and with 
a clock movement to revolve the 
chart once in one hour, but for con- 
tinuous daily records the standard 
twenty-four hour charts are recom- 

These instruments are manufac- 
tured by Wm. H. Bristol, 45 Vesey 
Street, New York City, who will fur- 
nish fuller information to anyone who 
is interested. 

New "Weston Instruments 

A new departure has been made by 
the Weston Electrical Instrument 
Company in the production of ac- 
curate and yet low-priced ammeters 
and voltmeters for switchboard work. 
The direct-current instruments are 
called the Eclipse type. They work on 
the "soft-iron" principle, but have 
been designed after years of investi- 
gation to eliminate the disadvantages 
ordinarily possessed by such instru- 
ments. The alternating-current in- 
struments work on similar lines. This 
is the first alternating-current am- 
meter produced by this company, it 
having been found impossible hereto- 
fore to construct such an instrument 
possessing sufficient accuracy. The 
new instruments are dead beat, very 
sensitive and are practically free from 
hysteresis error. They are made in 
two sizes, one for large and the other 
for small switchboards. The ammeters 
are made in sixteen ranges from one 
ampere to 500 amperes capacity. The 
voltmeters are made in seven ranges 
from 75 to 750 volts full scale de- 

mechanical and electrical improve- 
ments as a result of their experience 
during the last five years. 

The general form of the tubes of 
the several types of lamps remains 
the same, together with the quality and 
the high efficiency of the light. The 
improvements made in" the new 1908 
lamps are, for the most part, in the 
auxiliary and the method of installa- 
tion. The new design makes it possi- 
ble to handle each lamp outfit as a 

joint. Where it is necessary to hang 
the lamp at some distance from the 
ceiling, a pipe or conduit of the prop- 
er length can be used between the 
plate and the auxiliary. Where 
lamps are hung from outlet boxes the 
ceiling plate can be dispensed with. 
When desired, the lamp may be hung 
from a wire rope by screwing a hook 
to the insulating joint and properly 
guying the lamp so that it will not 



The 1908 Cooper-Hewitt Lamp 

The 1908 model of the Cooper- 
Hewitt lamp has a number of im- 
provements which add much to the 
commercial value of the lamp. The 
Cooper-Hewitt Electric Company, 
New York, are the sole manufacturers 
of mercury vapor lamps, and have in- 
corporated in the new model several 

unit, and has greatly simplified the 
hanging of the lamp. 

The parts and functions of the aux- 
iliary to the lamp tube remain the 
same, and the changes are along the 
lines of improvement in the manu- 
facture and the assembling of the 
auxiliary. The new casing of the 
auxiliary resembles, in a way, the 
housing of an arc, being cylindrical 
in form and about eight inches in di- 
ameter and 10 ins. high, with the 
supply lines entering binding posts at 
the top. The lamp rod, which sup- 
ports the tube, is attached directly to 
the auxiliary by a pivot screw. 

The hanging and connecting of the 
new lamp is extremely simple. A 
ceiling plate, a short-threaded pipe 
nipple and an insulating joint are 
furnished with each lamp outfit. 
These are fastened to the ceiling, and 
the auxiliary screwed to the insulating 

The Tungsten Electric Lamp_ Co. 

pleasure in printing the follow- 
ing communication from the 
newly organized "Tungsten Electric 
Lamp Co." : 
Editor Electric Age, 

45 East 42d St., New York City. 

Dear Sir: We, naturally, believe 
that improvements resulting in better- 
ing the conditions of all, or any por- 
tion of the people of this country, 
should be made known to them. 

One of the purposes of electrical 
magazines is probably to enlighten the 
readers and to keep them thoroughly 
informed regarding advancements 
taking place in the electrical field. 

By request of Mr. J. C. Fish, Presi- 
dent of the National Electric Lamp 
Association, we take pleasure in send- 
ing you information regarding The 
Tungsten Electric Lamp Company. 



January, 1908 

Our object is to aid in having all par- 
ties who are, or should be, interested 
in illumination derived through the 
medium of electric incandescent il- 
luminants realize the great advance- 
ment which will be made in the light- 
ing field by the Tungsten lamp. There 
is absolutely no doubt as to the suc- 
cess of this illuminant. The most re- 
liable lamp manufacturers in the 
country know this to be a fact and 
are financially interested. They are 
giving their close personal attention 
to The Tungsten Electric Lamp Com- 
pany, as well as having hired the best 
talent in the country to guide this new- 

The advent of the Tungsten lamp, 
we think, should be considered a mat- 
ter of as great importance as the pro- 
duction of the first electric incan- 
descent lamp, the Welsbach mantle, 

Yours very truly, 

Engineering Department. 

When considering electric incan- 
descent lamps, we must now include 
the tungsten. This new illuminant 
bids fair to supersede all others in 
many places. Its success is guaran- 
teed, if for no other reason than the 
fact that 16 of the most prominent 
lamp manufacturers in the country 
have seen fit to back their faith in it 
by organizing, under the laws of the 
State of Ohio, The Tungsten Electric 
Lamp Co., with a capital of $100,000. 
Following are the incorporators : 
Messrs. G. G. Lockwood. A. C. Garri- 
son, H. B. Vanzwoll, A. S. Terry. 
L. P. Sawver. H. H. Gearv. T. C. Fish, 
H. C. Rice, E. H. Haughton, E. \Y. 
Gillmer, J. B. Estabrook. W. D. Pak- 
card, Wm. Coale, E. J. Kulas. T. E. 
Randall and T. W. Freeh. Jr. 

Mr. Freeh was elected President 
and Manager. Mr. Randall is Vice- 
President and Supervising Engineer. 
The former has devoted many years 
to the manufacture of lamps, the lat- 
ter has been a lamp engineer since 
1886, when he filled that position for 
the Thompson-Houston Co. 

These two gentlemen have taken 
several trips across the Atlantic to 
thoroughly study the laboratory work 
being conducted in foreign countries 
on tungsten lamps, as well as to in- 
vestigate the methods used in manu- 
facturing them. 

Messrs. Freeh and Randall have 
held many consultations with the other 
members of the Company, and after 
much experimenting in improving 
upon the tungsten lamp as produced 
in other countries. As the carbon 
lamps in the United States are su- 
perior to those made in other lands, 
so is the tungsten. 

The factory of The Tungsten Elec- 

tric Lamp Company is located in 
Cleveland, O. It is 55 by 112 ft., four 
stories high, with a basement, and fire- 
proof throughout, concrete cement be- 
ing principally used in its construc- 
tion. The total floor area is 30,000 
sq. ft. The present capacity is not 
over 2,000,000 a year, but for several 
reasons can be increased within a few 

Tungstens are being made up in 40 
and 60 watt-units for multiple service 
and 40, 50 and 75 watt-sizes for street 
series work, the amperes on the last 
three being 4, 5.5, 6.6 and 7.5. The 
mean horiontal efficiency is 125 watts 
per c.p. 

The Tungsten light is extreme'v 
white, and can be used to advantage 
as a substitute for sunlight when the 
latter is desired, but cannot be ob- 


Belles Letters 

HE correspondence of any large 
company has a general same- 
ness, which is sometimes deaden- 
ing. One of our large manufacturers 
received the following, which departed 
from the usual and brought a smile 
to the faces of those who handled it : 

"Gents in refference to the 5 H P 
motor i have i had it repaired by a 
Plumber he soldered the Cylinder on 
each end and it works all right just 

"Resp. yours" 

The following letter came from a 
central-station manager with a dis- 
tinctive ability to put his words into 
form. It is seldom that letter writer- 
in business put a "call" into any other 
shape than an out and out complaint 
in the shortest possible phrasing. This 
letter received an attention which a 
dozen straight complaints would n< it 
have received and more than demon- 
strated that business is hard only on 
the surface. 

John Doe Mfg. Co. 

Gentlemen: We turned on the 100 
arcs of your company's make for the 
first time (for service) last evening. 
The 60 or more 50 c-p. incandescents 
had been in service when required 
for some nights, and were the sub- 
ject of general praise and UNQUAL- 
for the arcs — my dreams last night 
were troubled. I had not seen the 
bulletin of that particular lamp, nor 
been present at the sale, but immedi- 
ately upon knowing that opal inner 
globes had been sent, had demanded 
that they be replaced by clear ones, as 
specifications did not call for "dim- 
mers" of 75 per cent, efficiency or 

Your Mr. X had stood strongly for 
the opal globes, but said that they 

came on evervthing would be 
sun. As full 
on and they 

would be replaced by clear ones if 
not satisfactory on trial. We tried 
them, and, as I said, my dreams that 
night were troubled. 

I saw the lamps turned on, and 
as the nebulous spots of light appeared 
in places about the town there were 
no "Ahs" nor "Ohs" from a pleasantly 
surprised multitude, but many passers 
who HAPPENED to see one' of these 
lamps in a direction in which the moon 
never could be seen, rubbed his eyes 
and had sudden doubts of his own 
sanity as he saw what looked like a 
feeble resemblance of that luminary 
peeping at him as through a thick 

When they had collected their wits 
somewhat and began to discuss the 
matter with others, the idea became 
fairly general that what had been seen 
was 'PROBABLY the new arc lamps 
which were being tested by a very 
feeble current, and that when "full 

like unto the midda\ 
current was already 
looked in vain for any betterment, 
jeers, sneers and derisive laughter 
came from the crowd, and it was not 
until the abundant and mellow light 
of the 50 c-p. incandescents fell upon 
his pathway and upon his face that 
the impression was partially dispelled 
from the mind of the taxpayer that 
he had been sold and swindled, not 
knowing, poor soul, that the fault lay 
not in the innocent arc, but that he. 
the public, and even the manufacturers 
(when controlling spirits are SUP- 
POSED to hold things in balance with 
their COMMON-SENSE) were all 
the victims of the superabundant and 
"wise in his own conceit — expert kid," 
who had boxed up a large part of the 
light actually produced, in order that 
such as managed to escape might have 
better diffusion and a mellowness 
somewhat approaching the mellowness 
of the brains of the "expert." As the 
taxpayer went home we heard the in- 
candescent singing : 

"We're just old-fashioned lamps. 

But we will light your feet. 
That costly spark of many 'amps 

Which hangs upon the street. 
Calls itself an ark lamp 
Of a thousand "units" power; 

But when summer nights are damp 
You may see from out your bower 

Some tiny flashing lights 
By little bugs displayed, 

Which for volume and for bright- 

Would put those arc lamps in the 

I awoke sighing deeply, and could 
but say — Ah, beautiful incandescent! 
Absolutely truthful, as well as beau- 
tiful. Would that all lamps might 
emulate your worth. 

January, 1908 



The arcs ran one night. They are 
shut down to STAY down until 
CLEAR inner globes, arrive. One 
black eye is enough for either your 
company or myself. NO FACTORY 
tific reports," SIMPLY CLEAR 

Yours truly. 

Portable Standard Integrating 


Jan. 15, 1908. 
Editor of The Electrical Age, 
New York City. 

Dear Sir: In the article entitled 
"Portable Standard Integrating Watt- 
meter for Testing Alternating Current 
Service Integrating Meters," which 
appeared in your issue of December, 
1907, the following statement is made 
in the first column directly under the 
tabulation on page 476. 

"The per cent, of error in the meter 
under test may be found directly by 
dividing the number of revolutions of 
the standard by the number of disk 
revolutions made by the service 

In my opinion, this statement is 
erroneous, for the reason that the 
reading of the standard should be 
taken as 100 per cent, and the reading 
of the service meter should be ex- 
pressed in terms of the standard. For 
instance, in example 1, if the num- 
ber of revolutions made by the stand- 
ard is 1.03 and the revolutions of the 
service meter is one, then the per- 
centage of accuracy is 1/1.03 =0.971 
nearly, or the service meter is 2.09 per 
cent, slow, instead of three per cent, 
slow. Again, if the number of revolu- 
tions of the standard is 0.97 and the 
revolutions of the service meter is one, 
then the percentage of accuracy is 
1/0.97 — I -°3 I nearly, or the service 
meter is 3.1 per cent, fast instead of 
three per cent. fast. 

Similarly, I think that all the other 
calculations in the article and tabula- 
tion are wrong, due to the funda- 
mental error in expressing the per- 
centage accuracy of the meter under 

Believing that the writers of the ar- 
ticle would desire to make a correc- 
tion, I have taken the liberty of calling 
the error to your attention. 
Very truly yours, 

W. F. Howe. 

Schenectady, N. Y. 



Defendant was negligent in direct- 
ing an employee to tie an electric light 
wire to an improper insulator with a 

piece of common iron wire, where the 
whole danger might have been 
averted by having the current cut off 
from the light wire. Cumberland 
Telephone & Telegraph Co. v. Graves, 
Adm'x. Court of Appeals of Ken- 
tucky. 104 Southwestern, 356. 


One of the wires of an electric 
power company extending along a 
highway being down, so as to more or 
less obstruct the road, deceased, one 
of a pleasure party driving along the 
road, got a rope, threw it over and 
fastened it to a broken insulator on 
the wire, and while reaching for a pole 
on which to tie the rope, got so near 
the wire as to receive a fatal shock. 
Held, he was guilty of contributory 
negligence as matter of law. Shade 
v. Bay Counties Pozver Co. Supreme 
Court of California. 92 Pacific 62. 


An ordinarily bright and intelligent 
boy, twelve years old, living in a city 
in which electric light and power 
wires are in constant use on nearly all 
of the principal streets and highways, 
who, having knowledge of the danger, 
but not of its extent, purposely takes 
hold of such a wire in order to obtain 
a shock, and is injured therebv, is as 
a matter of law guilty of contributory 
negligence. Jolinston v. New Omaha 
Thomson-Houston Electric Light Co. 
Supreme Court of Nebraska. 113 
Northwestern, 526. 


An electric light company is not 
liable for injuries through electric 
shock caused by plaintiff driving into 
wires, which trespassing boys had 
thrown over a light wire, and which 
had just come into contact with the 
current where the insulation was off 
the light wire for about an inch. 
Lnehrmann v. Laclede Gaslight Co. 
St. Louis Court of Appeals, Missouri. 
104 Southwestern, 1128. 


Where insulation on certain electric 
wires, with which plaintiff came in 
contact and was burned, was not in- 
tended as a safety device, but was only 
for the protection of the wires, it was 
error to permit the jury to find that 
such insulation was old, weather-worn, 
broken, and out of repair, so as to 
afford no protection against the elec- 
tric current to a person coming in 
contact therewith. Rasmussenv. Wis- 
consin Traction, Light, Heat & Pozver 
Co. Supreme Court of Wisconsin. 
113 Northwestern, 453. 


Notice of the unsafe condition of an 
uninsulated electric light wire may be 
imputed to the light company from the 
fact that it has been in that condition 
for a considerable length of time, 
since thorough insulation is indispen- 
sable to confine the current to the wire, 
and the company's duty to keep 
its wires insulated is a continuing one, 
requiring careful and continuous in- 
spection. Luchrmann v. Laclede Gas- 
light Co. St. Louis Court of Appeals, 
Missouri. 104 Southwestern, 1128. 


Evidence that it was the custom of 
linemen to step on and make use of 
dead wires in climbing poles was ad- 
missible, notwithstanding a warning 
bulletin from the telephone company 
respecting poles having electric light 
wires attached to them, which related 
only to ordinary risks. Leque v. 
Madison Gas & Electric Co. Supreme 
Court of Wisconsin. 113 North- 
western, 946. 


An electric light company, furnish- 
ing and maintaining a lamp in a store 
and receiving monthly compensation 
for the service, must use reasonable 
care in placing it, and to that extent 
the company is a tenant in possession 
of the store as to all persons lawfully 
entering it. Fish v. Waverly Electric 
Light & Pozver Co. Court of Appeals 
of New York. 82 Northeastern, 150. 


A city owning and operating an 
electric light plant, as authorized by 
Rev. St. 1895, art. 421, empowering 
cities to provide for lighting the 
streets, etc., may, after discharging its 
duty to the public, sell its surplus elec- 
tricity to private citizens for lighting. 
Crouch v. City of McKinney. Court 
of Civil Appeals of Texas. 104 South- 
western, 518. 


"The Treatment of Belts and Ropes 
for Service and Profit" is the title of 
an attractive 88-page pamphlet which 
ought to be in the hands of every 
superintendent or engineer because of 
the information it gives on belts and 
belt management. It contains more 
information about the every-day 
practical uses of belts than anything 
with which we are acquainted, quite 
aside from its treatment of Cling-Sur- 
face. .Address Cling-Surface Co., 
Buffalo, N. Y. 



January, 1908 

Electrocraft Approved Electric Fit- 
tings and Revised National Electrical 
Code, 1907. Price, 50 cents. Electro- 
craft Publishing Co., Detroit, Mich. 

The Standard Steel Works, Phila- 
delphia, Pa., has sent out a handsome 
catalogue covering spiral and elliptical 

The Lazier Gas and Engine Co., 
Buffalo, N. Y., has sent out a unique 
and expensive catalogue of its product. 
It contains a blue print of each ma- 
chine and a description of each type 
with much useful information. 

General Electric Bulletin No. 4555 
describes the application of electricity 
to cement plants and gives a large 
amount of information regarding the 
different processes and the apparatus 

Fort Wayne Electric Bulletin No. 
1 102 gives a very lengthy list of plants 
operating direct-current, direct-con- 
nected generators of its manufacture. 
The company is also sending out its 
bulletin index for January. Bulletin 
No. 1 103 describes the series arc light- 
ing system. 

The Allis-Chalmers Co. is sending 
out a very handsome publication, en- 
titled "Keewatin Flour Mills," illus- 
trating some of the important mills 
built by the Allis-Chalmers Co. 

"Paiste Specialties" is the title of 
a 200-page catalogue now being 
mailed by H. T. Paiste Co., Philadel- 
phia, Pa. 

" Lifting Magnets," issued by the 
Electric Controller and Supply Co., 
Cleveland, Ohio, is a very handsome 
and complete engineering catalogue 
profusely illustrated in the engraver's 
best art. It is a complete text-book 
on the subject. 

Murray Corliss engines are illu- 
trated very completely in a catalogue 
now being sent out by the Murray 
Iron Works Co., Burlington, la. 

Westinghouse Bulletin No. 1137 
covers intergrating wattmeter for 
single-phase and polyphase alter- 
nating-current circuits and for direct- 
current circuits. 

Allis-Chalmers Bulletin No. 2027 
covers a description of the hydro- 
electric plant at Trinity River, Cal. 

name to Habirshaw Wire Co., as of 
January 1, 1908. 

The company was organized more 
than twenty years ago by Dr. Habir- 
shaw, who contemplated the building 
of deep sea cables, in which gutta 
percha was used extensively. His se- 
lection of the names India Rubber and 
Gutta-Percha recalls the fierce con- 
troversy of a past generation of en- 
gineers over the respective merits of 
insulating materials. The total elimi- 
nation of anything suggestive of the 
insulating compound in the present 
trade name is therefore significant. 

The Bureau of Supplies and Ac- 
counts, Navy Department, will enter- 
tain bids January 31st for the furnish- 
ing of miscellaneous electrical supplies 
at New York and one 5-ton electric 
truck at Washington, D. C. ; also for 
rolled bronze, soft-sheet brass and cop- 
per at Boston, Mass. 


The India Rubber and Gutta-Percha 
Insulating Co., New York, of which 
Dr. W. M. Habirshaw is president, 
and J. B. Olsen, sales manager, an- 
nounces the change of its corporate 


The Rail Joint Co., New York, 
eclipsed its best production record in 
the year 1907. As the exclusive 
maker of base-supporter rail joints, 
and the largest producer of rail joints 
in the United States, it has received 
contracts from every quarter of the 
globe. At the present time it is en- 
gaged on the Panama Railroad con- 
tract. From present indications, the 
outlook for business for the coming 
year is good, and the company is hope- 
ful of continuing its large output dur- 
ing 1908. 

The American Thread Co., Watup- 
pa, Mass., recently purchased a dupli- 
cate 1500 kw. Allis-Chalmers steam 
turbine ordered some months ago. 

Mr. Geo. R. Hall has formed an 
engineering partnership with John 
Crawford, at Boulder, Colo. Their 
first contract is the building of a 13,- 
200 volt three-phase transmission line 
about nine miles long, from Boulder 
to Salina. Colo., with the equipment 
of a substation for the distribution of 
300 h.p. to the mill and mine of the 
Pollock Mining and Milling Co. 

The Brush Electric Light and 
Power Co., of Galveston, Tex., has 
recently installed two Allis-Chalmers 
steam turbines of 1500 kw., normal 
rating. The steam plant, which con- 
sists of Heine boilers, has been in- 
creased by adding Erie and Stirling 
water-tube boilers. 

Voltax, a well-known anti-corrosive 
compound manufactured by the Elec- 
tric Cable Company, New York, has 
been specified for painting the Brook- 
lyn suspension bridge. 

Common Sense is the title of an 
attractive pamphlet issued monthly by 

the Electric Controller & Supply Co., 
Cleveland, O. It is interesting and 
well illustrated, and if the succeeding 
issues hold to the standard of the 
initial number. Common Sense will 
have a wide circulation. 

The United Verde Copper Co. has 
installed a second three-motor North- 
ern traveling crane, in addition to the 
50-ton crane supplied several months 

The Lord Electric Co., New York, 
has concluded arrangements with 
Chas. I. Earll for the manufacture and 
sale of his trolley retriever and catch- 
er. It is a most ingenious contri- 
vance, very strong and durable, easy 
of access, free from complications, at- 
tractive in structure and built on good 
engineering principles. Many of the 
largest of the most representative 
properties in the United States have 
adopted this device. 

W. R. Garton, formerly president of 
the W. R. Garton Co., Chicago, is 
now general manager of the Lord 
Electric Co., New York. 

The Standard Steel Works Co., 
Philadelphia, Pa., is sending out its 
catalogue on springs to those inter- 

January issue of Light, published 
by the United Electric Light & Power 
Co., New York, notes the lighting of 
the entire Manhattan approach of the 
207th Street bridge over the Harlem 
River by the company's mains and the 
replacement of the steam-driven draw- 
bridge by two 20-h.p. alternating-cur- 
rent motors. 

In the new Fulton Terminal Build- 
ings, New York, which will be ready 
for occupancy in May, 1908, the Ma- 
chinery Club of the City of New York 
will occupy the 20th and 21st floors, 
including a roof-garden. The mem- 
bership of the Machinery Club is 
limited by the present constitution and 
by-laws to a possible membership of 
750 resident, 500 suburban and 1000 
non-resident members. The member- 
ship list is already representative of 
the machinery interests of the United 
States. It is expected that a waiting 
list will have to be shortly established. 

On the 29th of January, 1908, 
Arthur Killyan, Esq., City of Niagara 
Falls, N. Y., will act as Referee 
in the petition of the Niagara Ta- 
chometer and Instrument Co. to be 
declared bankrupt. The petition is 
signed by Henry A. Francis, de Lan- 
cey Rankine, Chas. R. Huntley, Geo. 
J. Howard, Wm. A. Brackenridge, 
Benjamin V. Norton. Henry A. Fran- 
cis has been appointed temporary re- 
ceiver for the company. 


Volume XXXIX. Number 2. 
$ 1 .00 a year ; 1 5 cents a copy 

New York, February, 1 908 

The Electrical Age Co. 
New York and London 

Published monthly by The Electrical Age Co., 45 E. 42d Street, New York. 

J. H. SMITH. Pre*. C. A. HOPE. Sec. and Treas. 


Telephone No. 6498 38th. Private branch exchange connecting all departments. 
Cable Address — Revolvable, New York. 


United States and Mexico, SI. 00: 

Canada, $1.50. To Other Countries, $2.50 


Insertion of new advertisements or changes of copy cannot be guaranteed (or the following 
issue if received later than the 15th of each month. 


Business and Editorial Departments, 33 Bedford Street 

Strand. W. C. 

London Telephone, Post Office, 3807 Central. Private branch exchange. London Cable 

Address Revolvable, London. 

All remittances to London Office should be crossed London and County Banking Company. 


Subscriptions may be sent to the following branch offices or agencies, where trey will re- 
ceive the same careful attention as at the home office in New York : 

124, Queen St., Melbourne 33, Loop St., Cape Town 

359, George St., Sydney Unter den Linden, 5. Berlin 

61, Main St-, Yokohama Nevsky Prospect, St. Petersburg 

23, Esplanada Rd., Bombay 31, bis rue de Fauburg Monlmartre, Paris 

General Agents for United States and Canada : The American News Company 

Water Powers of the Southern 
-Appalachian Streams. 

The streams whose headwaters lie 
among the peaks of the Southern 
Appalachians, flowing westward to 
the Mississippi or eastward to the 
Atlantic, furnish opportunities for the 
development of water power so won- 
derful that the meagerness of their 
present use for this purpose is little 
less than marvelous. The position of 
these rivers as prime factors in the 
industrial growth of the South is well 
recognized, and the power develop- 
ment that has taken place in the re- 
gion thus watered is unparalleled in 
any other portion of the United 

Engineers of the United States 
Geological Survey, after making a 
careful study of the streams, the 
quantity of water they carry, and 
their fall in various portions of their 
courses, have estimated that they 
afford a minimum of about 2,800,000 
horse power, at least 50 per cent, of 
which, or 1,400,000 h.p., is avail- 
able for economic development. 
These figures, it should be noted, 
represent the minimum horse power. 
If auxiliary power were provided to 
supplement the water in short seasons 
of deficiency, 2]A times this amount 
might be profitably utilized ; and if 
the flood waters could be stored and 
the flow of the streams properly regu- 
lated, the minimum power available 
for economic development might be 
increased from three to 15 times. It 
is evident, therefore, that an estimate 
of the present value of these water 
powers, based on 50 per cent, of the 
minimum indicated horse power, has 
so many factors of safety that it is 
very conservative. 

An extremely low average of the 
present rental value of water power is 
$20 per horse power per year. The 
rental value of 1.400.000 h.p. would 
therefore amount to, 
which is equivalent to an income 

of 3 per cent, on a capital of S933,- 
000,000. But the resource represented 
by this water power is far greater 
than its present market value. The 
enormous and wasteful consumption 
of fuel is so rapidly depleting the 
supply that it must become more and 
more costly. As a result the present 
disparity between the cost of fuel 
power and that of water power will 
constantly increase, and the demand 
for water power will increase accord- 

The Decher Cell 

"One is at a loss to point out any 
radical or even important advance 
in primary batteries since the time 
of these fundamental improvements."' 
declared F. B. Crocker, past presi- 
dent of the A. I. E. E. and 
present professor of electrical en- 
gineering at Columbia University, in 
outlining the early development of the 
primary battery in a lecture before the 
American Electrochemical Societv. 
Oct. 8. 1906. on the Decker primary 
battery. In making this unqualified 
>tatement there would seem to be a 
deliberate evasion of the fact that 
practically none of the types earlier 
alluded to in the address are in use. 
and that the only battery now in use 
in any number for supplying power 
continuously is the Edison-Lalande 
type. So epochal was the develop- 
ment of the Edison cell among pri- 
mary batteries that it practicallv 
seized the commercial market. 

It is claimed for the Decker bat- 
tery that the cost of producing power 
is at the rate of 35 cents per h.p.h.. 
or 47 cents per kilowatt-hour, this 
calculation being a theoretical one 
based on the consumption of zinc and 
chemicals alone, the estimated figure 
reducing to 15 cents per h.p.h." or 
20 cents per kilowatt-hour where the 
cells are used in so great numbers as 
to justify reclaiming the zinc from the 
exhausted batterv thud by electrolvtic 

action. Exact data on this pr< 
have not been given forth, so that it 
is hardly possible to examine them. 
It is not exactly clear whether it is 
proposed to use Decker current in 
this process, in which case we should 
have the signal operation of decom- 
posing electrolytes by the same chem- 
ical forces which unite them in pro- 
ducing current flow and which we are 
taught is impossible under the laws of 
electrolytes, being a chemical process 
very like perpetual motion in mechan- 
ics ; or whether the current for re- 
claiming the battery solution will be 
central station power service at two 
or three cents per kilowatt-hour. If 
the latter, we are of the opinion that 
the operation would have to be on a 
very large scale to be successful. The 
cost of electrolvtic current can be but 
a sniall part of the cost of such a 

It is admitted that 47 cents per kilo- 
watt-hour is a high figure, but never- 
theless it is ambitiously proposed to 
enter the fields of light, power and 
vehicle work with the Decker battery, 
and to do SO an enormous sum of 
money is to be raised. The promoters 
of the company of which E. B. Crock- 
er is president, are endeavoring to raise 
$3,000,000 by selling 200.000 shares 
of stock, par value. 100 at $15 per 
share, making the total capitalization 
S20.000.000. How large these figures 
are in the primary batterv field, and 
how certain the promoters feel of 
their ground ma}- he gauged by the 
fact that the entire amount of money 
now invested in the manufacture of 
primary batteries in the United States 
is somewhat less than Si. 000,000, or 
one-third the amount which is ad- 
judged necessary to demonstrate the 
revolutionary changes which the new 
cell i-^ destined to make. 

The Decker Electrical Mfg. Com- 
pany, which is the name oi the con- 
cern organized to float the battery, has 
offices at Baltimore. Philadelphia, 
Xew York and main offices at Wil- 




February, 1908 

mington, Del., at all of which it has 
been selling stuck during the last year. 
A factory has been built at Phila- 
delphia and batteries are being manu- 
factured for sale. 

The battery is. in essential elements. 
not new. It is the old bichromate 
cell. The novel thing about it is the 
mechanical arrangement for filling 
and emptying the cell, which may be 
accomplished by a siphon action or by 
air pressure. This arrangement is a 
welcome improvement on the old 
plunge type of bichromate cell of 
which this may be described as a new 
and mechanically improved type. 

in Prof. CrockerV address, he states 
that the new type gives 14.7 watt-hr. 
per lb. of battery. This figure was 
obtained only after shaking the bat- 
tery to get a further output by facili- 
tating the action of the depolarizing 
bichromate solution, which figure is 
further stated significantly to be about 
'"twice as great as that obtained in 
standard types." In this statement 
the writer is evidently in error, as 
there are cells on the market giving 
nine to 14 watts per lb., and a new 
storage cell has this figure just about 
d< mbled in its claims. 

There is evidently, also, a differ- 
ence of opinion as to the watt-hour 
per lb. output of the battery, as Prof. 
Carl Hering. after a test of a 16-cell 
battery, stated in a lecture before the 
American Electrochemical Society. 
( >ct. 18. 1906, "that the capacity was 
nearly at the rate of 10 watts per lb. 
of complete cell." 

It is seriously proposed to employ 
the Decker battery in train lighting, 
and it is stated that the cost per car is 
S4 per night with tungsten lamps as 
against So per night with carbon 
lamps and storage batteries. Since 
the latter figure on storage batteries 
will, on the basis of tungsten effi- 
ciency, be only one-third as much as 
carbon filament lamps, the true com- 
parison figure is $2 per night for 
storage batteries, or half as much as 
the cost of the Decker outfit. It 
would seem, if these figures are cor- 
rect, that in the present form of the 
proposition this is not a promising 
field for the exploitation of the cell. 

It would hardly appear that a bet- 
ter case could be made out in electric 
vehicle work for the new cell. To 
mention one item alone the discharge 
time for the cell is too short. It is 
given as 5'^ hr.. which is too short 
an interval for a trucking day of 
8 to 12 hr. 

It would be necessary with the 
Decker outfit to employ two storage 
tanks, one for chemicals to renew the 
battery fluid and one to provide stor- 
age for the spent liquids, together 
with an automatic pump to accomplish 
the transfer of liquids to and from 

the battery. A reasonable amount of 
piping and drip cocks would be neces- 
sary to arrange all of this. So in 
place of the simple charging process 
of the storage batter}- we should have 
a double reserv< >ir storage to accom- 
plish at the garage, and a renewal of 
all the zinc elements every day or 
other day — present zincs last for three 
charges only. 

The replacement of the zinc in itself 
would be a serious task in a battery, 
which would have to contain at least 
70 zincs to get potential. Vehicle mo- 
tors are no- volt and an average E. 
M. F. of the Decker cell as stated by 
Dr. Hering is 1.64 volts. 

If in addition to the labor items 
mentioned above it were necessary to 
operate a reclaiming plant, it can 
easily be seen that the outfit cannot 
be so easily or cheaply bandied as 
sti >rage battery charging. 

If now we take up the question of 
cost operation we are still further 
surprised that the Decker cell could be 
seriously considered for vehicle work. 

In the Edison type of storage bat- 
tery it co^ts about -$i per day for cur- 
rent at six cents per kilowatt-hour, 
practically nothing for fluid renewal, 
and only from 15 to 20 cents a day 
for maintenance. The total is about 
Si. 25 per day. The corresponding 
figure for the Exide cell is probably 
S_> per day at a minimum. Both of 
these values are attained in electrical 
vehicle operation. 

Let us now compute the cost of 
Decker current. Assuming the fig- 
ures which have been given by Prof. 
Crocker, the president of the com- 
pany, and they may be taken as au- 
thoritative and at the same time as 
the most favorable which can be 
shown for it. 

Assuming an average output for 
the vehicle motor of 1.75 h.p.. an 
average time of running of eight 
hours, and we have a current require- 
ment of 14 or 10 kw-br. If the 
average efficiency of the motor and its 
resistances C 70 per cent., the battery 
imput will have to be 14 kw-hr. and 
at 47 cents per kilowatt-hour the total 
daily cost would be $6.58. Assuming 
a figure of 20 cents per kilowatt-hour 
to be attained by the reclaiming proc- 
ess, and this figure is not established 
by published tests. Decker current 
alone would cost $2.80 per day. Both 
of these current figures are greatly in 
excess of machine generated current, 
which can reach a value as low as 30 
cents per day with current at two 
cents per kilowatt-hour. 

The computations given above for 
the Decker outfit are probably too low 
for the reason that the 47-cent and 
the 20-cent rate are based merelv on 
the theoretical cost of battery mate- 
rials and do not include inspection, 

attendance, etc.. which are included 
in the storage battery figures given 

We are of the opinion that while 
the battery is destined for no exten- 
sive use for vehicle or lighting work, 
it may be found useful where there 
is no supply of service current. This 
field is widening daily, but on the 
whole, it is rather inconsequential. 
owing to the rapid spread of central 
power service. 

It is our opinion, also, that it were 
better for the men interested in the 
company to try the battery out on a 
reasonable business scale of operation 
instead of attempting to raise from 
the dear public a sum many times over 
that invested in the battery business 
at the present time. And while we 
are not surprised that a battery in- 
ventor sfionld take this means of rais- 
ing money, we are most emphatically 
surprised that an honored past presi- 
dent of the American Institute of 
Electrical Engineers, who is reputed 
to be heavily interested in and is 
known as one of the founders of the 
large Crocker- W heeler Co.. and is at 
the present time professor of electrical 
engineering at Columbia University — 
we are surprised that he should head 
a promoting venture of this sort. 
There is plenty of capital in the elec- 
trical industry itself in the control of 
men who know the commercial field 
for inventions without asking the un- 
knowing public for it. 


Mica is the most expensive raw 
product used in the manufacture of 
electrical machinery, and the economic- 
al employment of it is rather a difficult 
problem for the manufacturer. There 
is a choice of various sized pieces and a 
choice of various varieties of mica and 
the price ranges from $5 to $6 per 
pound for first-grade selected mica 
sheets, approximately 6 by 8 in. in 
size, down to uncut mica in 2-in. pieces 
at 13 cents per pound. North Caro- 
lina mica costs 25.7 cents, whereas 
Xew Hampshire mica costs 3.6 cents 
and Virginia 2.0 cents. 

As a single illustration, consider the 
matter of commutator insulation. 

For insulation between segments of 
commutators there is nothing superior 
to the Canadian amber mica and cer- 
tain varieties of India mica. The lat- 
ter is of about the same hardness as 
the hard-drawn copper segments of 
the commutator, and the whole wears 
down evenly. Considerable museovite. 
or white mica, is employed for this 
purpose, but its inferiority does not 
show up for some time. Black- 
specked mica, in which specking is due 
to minute crystals of iron oxid. is 
purchased in large quantities by elec- 
trical dealers. It is. however, danger- 

February, 1908 



ous to use it for insulation against 
high potential currents. It is said that 
it can be used for insulation purposes 
where the current is under a iooo 
volts. The temptation to use it can 
be gaged from the fact that it costs 
only from one-half to two-thirds as 
much as clear mica. Many of the 
large electrical manufacturers will not 
use this variety of mica under any 

According to the last report of the 
United States Geological Survey, the 
production of sheet mica amounted to 
1,423,100 lb. in the United States, 
of which about two-thirds was mined 
in North Carolina. In the year 1906 
about 3,000,000 lb. of mica was im- 
ported. Less than one-third of the 
entire quantity used in the United 
States in 1906 was of home produc- 
tion. The increase in the year 1906 
over the year 1905 of sheet mica was 
about 78 per cent., the total consump- 
tion reaching about 4,560.000 lb. 

The Professional Electrical 

The teaching of electrical engineer- 
ing is still very far behind its sister 
professions and is in a generally un- 
satisfactory condition. The frequent 
discussion of the topic itself, the una- 
nimity of agreement among those who 
are most interested, and the constant 
discussion of the matter by the manu- 
facturing engineers themselves, all too 
clearly point to its unsatisfactory state. 

Most of the eminent critics have 
contented themselves by pointing out 
defects in the present pattern of in- 
struction, some few have suggested 
that none but practical men should 
teach, and lest these should go to seed 
it has been urged that the incumbent 
of the professorial chair and his as- 
sistants should at year-long intervals, 
take a dip back into commercial en- 
gineering. But we do not recall one 
genuine effort to get at an understand- 
ing of the matter. 

As in a problem of science it be- 
comes necessary to establish limiting 
conditions and controlling influences, 
so in an effort to make a satisfactory 
curriculum it should be determined, 
first of all, to just what point instruc- 
tion will be carried, where practice 
shall begin and how much of it shall 
be given. 

Of the four years of study in the 
usual electrical curriculum three are 
spent in assimilating the rudiments of 
raw science and in laying the ground 
work for later instruction. One short 
year is devoted strictly to electrical 
engineering. It is not surprising that 
graduates have to spend practically 

two or more years in apprentice work 
in order to fit them for professional 
work. No other professional school 
attempts to teach its novitiates in so 
short a time and electrical engineering 
is not less complex than any other 
branch of professional work. 

More time must be given to elec- 
trical engineering per se, and the 
amount and character of this work 
must be approximately specified, so 
that it joins to the present-day re- 
quirements of electrical engineering. 

Let the A. I. E. E. appoint a com- 
mittee of three from manufacturing 
companies, three from operating com- 
panies, three consulting engineers 
and three members selected from en- 
gineering schools. Such a committee 
should know what is best. Let them 
mark out what attainments a student 
should have who knocks for admit- 
tance to practical work. Let them fix 
the plan. Then suit the educational 
work to the plan and trim all the sub- 
sidiary instruction to the main lines 
of professional teaching. We are of 
the opinion that such a committee, in 
its recommendations, would probably 
reverse the present apportionment of 
time between raw science and elec- 
trical engineering. One result is al- 
most certain. Those institutions 
which could adjust themselves to a 
new alignment in their professional 
courses would do so. The time has 
arrived when the study of electrical 
engineering should be dignified by the 
creation of professional schools. 

ficial, in a sense, the embarrassment 
of the moment was seriously as it- 
pinch was felt temporarily"' — what- 
ever that may mean. 

The "Westing'House Situation 

One of our esteemed but fallible 
contemporaries, which had nothing to 
say in its strictly editorial column in 
the week following the announcement 
of the W'estinghouse receivership, 
now announces that it is a matter of 
pleasure to note how rapidly the 
W'estinghouse situation is righting it- 

This magazine last November an- 
nounced confidently that a situation 
of so vital importance to the entire 
electrical industry would right itself. 

Succeeding events are proving its 
estimate of the situation to he correct. 

The W'estinghouse Lamp Company 
has been discharged from its receiver- 
ship, and it is now announced that 
99 per cent, of the $7,000,000 liabili- 
ties of the W'estinghouse Machine 
Co. have assented to the proposed 
financing plan for that company. 

The great courage and indomitable 
will of George W'estinghouse are once 
more riding over a situation of which 
our contemporary now says, "As one 
looks back the vista to last ( )ctober 
shortens up. and it is seen how super- 

Trie Tungsten Filament 

With the advent of the new tung- 
sten filament we are likely to see a 
revival of the old struggle to get a 
sufficient downward distribution from 
the lamp itself. It will be recalled that 
the old Edison hair-pin filament gave 
a downward distribution of only 2 c-p. 
on a 16-c-p. lamp. The 2 c-p. of 
downward distribution just about rep- 
resented the proportional part of the 
loop illumination. When the double 
hair-pin filament was introduced the 
downward distribution was doubled, 
and with the introduction of the 
double oval filament the light was 
nearly quadrupled. In order to in- 
crease the downward illumination of 
this type of lamp common use is made 
of prismatic reflectors capable of in- 
creasing the downward illumination 
ordinarily to about 40 c-p. The exact 
intensity of illumination can be varied, 
however, by the design of the reflect- 
ing bowl. 

While the prismatic bowl is not 
ordinarily required in the common 
form of carbon lamp giving 7 c-p. 
downward from a 16-c-p. lamp, it is 
almost an absolute necessity with the 
tungsten lamp giving only about 8 c-p. 
downward from a 40-c-p. lamp. 

In the new tungsten filament we 
have no portion of the filament radia- 
ting directly downward. Consequently, 
it gives the old familiar black spot of 
the hair-pin filament, and is rather un- 
suited for use in individual lights. 

The Step Bearing' 

It is a question of considerable mo- 
ment in view of the large number of 
vertical type turbines in use whether 
some means ought not to be employed 
to provide an emergency source of oil 
pressure for the step bearing. The 
pressure at this point is so enormous 
— f>oo to 850 lb. per sq. in. — that the 
slightest failure of the pump is liable 
to bring the machine down on its 
step and put it out of action. If this 
were merely a momentary affair like 
the blowing of a fuse or the tripping 
of a circuit-breaker it were unimpor- 
tant. But it is more serious. The 
machine is put out of commission for 
a considerable length of time and the 
bearing is usually destroyed. While 
it is not an expensive matter to re- 
pair the damage, the unreliability of 
the device i> seriously a matter for 
consideration. Sound engineering is 
usually reliable engineering. 

The Central Station Distributing System 

Systems of Distribution 

THE early development of elec- 
trical properties having been 
limited to the use of the direct- 
current dynamo and motor, direct-cur- 
rent systems were the pioneers in cen- 
tral-station distribution work. 

Electricity was generated and trans- 
mitted at the voltages at which it was 
delivered to the consumer, and sta- 
tions were located with reference to 
the location of the load as far as ] 

These developments were begun in 
the larger cities, where the demand 
for electricity was greatest and where 
the geographical situation was such 
that the use of electricity was confined 
within comparatively small areas. 
This led naturally to the development 
of interconnected networks of low- 
tension mains, supplied by feeders at 
various points. These networks were 
interconnected through junction boxes 
at intersecting points, the lines from 
which customers' services were taken 
being called mains and lines from the 
station to definite points in the system 
being known as feeders. 

Having had an early start, and hav- 
ing been fully justified by the demand 
for direct current for elevators and 
other variable speed machinery, these 
direct-current systems are found in 
the central portions of many of the 
larger cities of the United States and 

The extension of distributing sys- 
tems into the more scattered portions 
of the larger cities and the necessity 
of reduction of the cost of installation 
directed attention to systems in which 
higher voltages could be employed. 
The possibility of the use of the trans- 
former naturally led to the use of al- 
ternating current generated at a volt- 
age of about noo and transformed 
down to no at the consumer's prem- 

This made possible the installation 
of a distributing system covering the 
more remote parts of a city with a 
much less weight of copper in dis- 
tributing mains and feeders. These 
earlier installations were operated 
single-phase at a frequency of 125 to 
133 cycles, which permitted a less ex- 
pensive design of transformer, but was 
not well adapted to motor service. 
The growth of the demand for power 
service resulted in the development of 
polyphase systems of 60 cycles and 



Commonwealth Edison Co., Chicago. 

2200 volts, which permitted further 
economy of copper, and put the 
alternating-current stations more near- 
ly on a par with the direct-current 
systems in the matter of power service. 

The direct-current station managers 
were not slow to recognize the pos- 
sibilities of a combination of the poly- 
phase system with rotary converters 
for the purpose of transmitting large 
loads over considerable distances. 
This enabled them to consolidate their 
--mailer generating stations into one 
large plant, utilizing the smaller gen- 
erating stations as converter sub- 
stations. The introduction of the 
rotary converter made desirable a 
still lower frequency than 60 cycles. 
and 25 cycle generators were accord- 
ingly provided for the large stations. 
\ oltages were selected with less uni- 
formity, however, the range being 
from 6600 to 13,200. depending upon 
local conditions. 

\s a nsult of this evolution, the 
following general types of distributing 
systems have come into more or less 
extensive use : 

1. Direct current, two-wire and 

This is the simplest system of dis- 
tribution, since the electric current 
ses directly from the generator to 
the consumer without any intervening 
transforming or regulating apparatus. 
The voltage at the consumer's premi- 
ses is that at which the electricity is 
generated less the loss of transmis- 

The voltage is therefore limited to 
that at which the consumer may con- 
veniently utilize the current, which. 
for lighting purposes, may not exceed 
220 to 250 and for power purposes, 
500 to 600. The earliest incandescent 
lamps were limited to about no volts. 
and two-wire no-volt plants were the 
pioneers. The cost of feeders s> 
led Edison to devise his three-wire 
system in which two generators were 
connected in series, a third or neutral 
wire being carried from the midpoint 
between the generators. This per- 
mitted the use of 110- volt lamps con- 
nected between the neutral and either 
outer conductor, while motors could 
be connected across the outside and 
operated at 220 volts. This was 
equivalent to doubling the voltage of 
the system, as the load could be kept 
approximately balanced between the 

two sides of the three-wire system. 
The current for a given load was thus 
halved and the radius of transmission 
was doubled. At 220 volts the loss on 
any feeder is about 10 per cent, when 
it is carrying one ampere per 1000 
cir. mils a distance of 1000 ft. 
When a feeder on a 220-volt system 
is much over a 1000 ft. long it is 
therefore necessary to use more cop- 
per than is required for safe current 
carrying capacity, or to install a 
booster to make up the excessive loss 
1 'ii the longer feeder. 

The location of a 110-220 volt two- 
wire or three- wire station or sub- 
station should therefore be such that 
the average length of the low-tension 
feeders does not exceed 1000 ft. and 
the longest feeder should not exceed 
1 to 3000 ft. In a 500- volt two- 
wire power system or in a 220-440- 
volt three-wire system the radius 
should be 2000 ft. for the average 
feeder, and not to exceed 5000 or 
6000 ft. for the longest. 

The no-volt system has so narrow 
a range that it is not advisable to use 
it for distribution purposes, except in 
small isolated plants. 

The three-wire no-220-volt system 
is suitable where the load is mostly 
within a radius of 1200 to 1500 ft. 
from the station and where there is a 
demand for direct current for power 

The 220-440-volt system may be 
used where the load is within a radius 
of half a mile from the station. This 
system is subject, however, to the 
serious handicap of low efficiency in- 
candescent lamps and the necessity of 
using 220-volt arc lamps, fan motors 
and similar small apparatus, which 
is more practical at no volts. 

The 500-volt system for power pur- 
poses is suitable where, the load is 
within a radius of about 5000 ft. and 
is found useful as a means of furnish- 
ing power for elevators and other vari- 
able speed machinery in cities where 
the existing alternating-lighting sys- 
tem is not well adapted to power work. 
This arrangement, how ever, requires a 
separate distributing system parallel- 
ing the lighting system which occupies 
pole and duct space and increases the 
ci 'st of maintenance. 

The foregoing limitations of direct- 
current systems therefore restrict the 
use of such svstems to localities where 

February, 1908 



power is required for variable speed 
machinery and where the lighting load 
is densely distributed near the point 
of supply. This is the condition in 
parts of large cities, in large industrial 
manufacturing plants and in some 
large amusement parks. 

2. Single-phase Alternating Cur- 

In this system alternating current 
is usually generated at about 2200 
volts and 60 cycles. It is delivered to 
feeders at the same voltage, and from 
the feeders distributed over primary 
mains to transformers located near the 
consumer who is to be served. These 
transformers reduce the pressure to 
the usual working voltages of 110-220, 

ing in copper incident to the use of 
the higher voltage. 

The single-phase system is not well 
adapted to locations where there is a 
considerable power load, owing to the 
limitations of the single-phase motor. 
It is best suited to cities where the 
power load is in units of not more 
than 20 to 25 h.p., and where no con- 
siderable saving would be realized 
from three-phase transmission. 

3. Two-phase system. 

In this system electricity is delivered 
at about 2200 volts by four-wire two- 
phase, three-wire two-phase or two- 
wire single-phase feeders and primary 
mains to transformers. These reduce 
the pressure to the usual voltages of 


Feedck Center 

■ ^<*at r fi>iAS£ Haws 

FIG. I. 

at which pressure the electricity is dis- 
tributed over secondary mains to the 
consumer's premises. 

The general arrangement of a feed- 
er and its primary mains on this sys- 
tem is shown in Fig. 1. 

This system being operated at 2200 
volts permits of distribution within a 
radius of two miles from the point 
of supply without the use of more 
copper than is required for current 
carrying capacity. More distant points 
may be reached by the use of booster 
transformers or potential regulators. 

The distributing mains also have a 
radius from the feeder end about ten 
times greater than that of a 220-volt 
system, which makes it possible to dis- 
tribute a widely scattered load from a 
smaller number of feeders, carrying 
heavier loads than would be feasible 
with low-tension feeders. 

The use of transformers and sec- 
ondary mains partially offsets the sav- 

110-220, or in some cases to 440 volts 
for motors only. 

Single-phase feeders are run in ter- 
ritory where no larger power service 

Two Phase 


Thmlil Phase. 


cost of two small transformers as 
compared with one large one. 

When the generators or transform- 
ers supplying a two-phase system may 
be L connected, as shown in Fig. 2, one 
wire of each phase may be made a 
Common or neutral wire and the feeder 
and main system reduced to a three- 
wire basis. 

The neutral wire in such a system 
carries the resultant of the current in 
the two phases, which i> 41.4 per cent. 
more than that in the phase wires. 
That is, in a feeder carrying 100 
amperes on each phase wire the neu- 
tral wire carries 141. 4 amperes. 

If the same size of wire is used on 
each leg of the circuit, the energy loss 
is the same as it would be in a four- 
wire feeder under similar conditions. 
There being but three wires, it is obvi- 



1 I 

Three. W/ntz. Two Phase. SrjrE/f 
FIG. 2. 

ous that there is but 75 per cent, as 
much copper required for this system 
as for a four-wire system under equiv- 
alent conditions. 

In cases where feeders are short 
and the carrying capacity of wires 
must be considered as well as regula- 
tion, it is sometimes necessary to use 

Two Phase. 

Three Phase. 

Two Phase 

FIG. 3. 

is required. Consumers requiring less 
than 5 h.p. are usually not given 
two-phase service, owing to the extra 

a larger conductor on the neutral, in 
which case the saving may not be more 
than 10 or 15 per cent. 



February, 1908 

In the primary distributing mains, 
where for mechanical reasons no wire 
smaller than No. 6 should be used, a 
saving of 25 per cent, is generally 

Where the system is used with a 
transmission beyond the limits of the 
generated voltage, it is usual to step 
up the voltage by two transformers 
connected as in Fig. 3. This is some- 
times known as the "Scott con- 
nection," and produces three-phase 
current on the secondary side, permit- 
ting the transmission to be made on 
the three-phase system. The reverse 
arrangement is usually used at the re- 
mote end to secure the simplicity of 
the two-phase system in distribution, 
but the distribution may be made on 
the three-phase system if the load is 
maintained in approximate balance on 
the three phases at all times. 

and where the lighting load is not 

The three-phase feeders may carry 
lighting on one phase only, as in the 
case where the load is mostly lighting 
and does not exceed 200 kw., or they 
may carry lighting on all phases, as is 
usual where load of 200 to 500 kw. 
come within an area which can be 
properly served by a single feeder. 
The primary mains are three wire 
where power or heavy lighting loads 
are to be served and two wire where a 
small lighting load only is carried. 

"Where three-wire feeders approxi- 
mately balanced, can be used on this 
system there is a saving of 25 per 
cent, in feeder copper over the single- 
phase system. Pressure regulation, 
however, is difficult if the load is un- 
balanced, since the adjustment of 
either potential regulator affects the 





■3>*C*.£ &1X3E Aff^gJ 

FIG. 4- 

The four-wire two-phase system is 
illustrated in Fig. 4. 

4. Three-phase three-wire system. 

Electricity is usually distributed in 
this system at 2200 volts and 60 cycles 
by means of two-wire single-phase or 
three-wire three-phase feeders and 
primary mains at the same voltage. 
This is stepped down by transformers 
to supply no-volt two-wire, 110-220 
volts three-wire single-phase, or 115- 
200 volt four-wire three-phase second- 
ary mains. 

The two-wire feeders supply terri- 
tory in which power is not required 

pressure on two phases and regulation 
is therefore somewhat cumbersome. 

Power services demanding 5 to 20 
h.p. may be supplied from two phases 
by the open delta connection as shown 
in Fig. 5, thus effecting a saving in 
transformer investment. 

The development of new power ter- 
ritorv or additional lisrhtinsr feeder ca- 
pacity is readily accomplished by the 
addition of a third wire to a single- 
phase feeder, thus converting it into a 
three-phase feeder. 

The arrangement of feeders and 
mains in this system is shown in Fig. 

5. Three-phase four-wire system. 

Alternating current of 60 cycles is 
supplied at 2200 volts between either 
phase and neutral point of a Y-con- 
nected generator or transformers and 
3800 volts between phase wires. Two- 
wire single-phase feeders and mains 
supply sections where no large power 
is required, being connected from 
phase to neutral at 2200 volts. 

Where the load in a given section 
consists of both lighting and power a 
four-wire three-phase feeder is run to 
a center of distribution so located that 
a proper arrangement of lighting 
mains may be made. If the lighting is 
not sufficient to load the three phases 
it may all be put on one or two phases, 
thus saving the expense of regulators 
and instruments until they are re- 
quired on the other phases. 

In cases where the lighting is suf- 
ficient to load three phases, but power 
load is very small, it is preferable to 

-tZOO -1 -ZZOO -» 

-3QOO -* 

ZZOO 2200. 

1-220 -m-ZZO -t 


I-220 ->■ ZZO -> 


Three. Wire. Three Phase. fou/t Wire, Three Phasc^ 

Of£/V DE.L.-TA. Co/VH£CT/OfV. 

FIG. 5. 

establish separate single-phase centers 
of distribution with single-phase 
mains. The lighting in a given section 
may thus all be carried from one phase 
and two-wire mains will suffice instead 
of the three- or four-wire mains, 
which are necessary where both light 
and power are to be carried. The extra 
power phases are carried only where 
they are needed for the few power 
consumers. In such situations the 
four-wire feeder permits three-phase 
transmission to be made from the sta- 
tion to the point where the single- 
phase feeders diverge, thus securing 
practically the same economy in cop- 

February, 1908 



per as in the case of a purely three- 
phase arrangement. 

The preservation of a balance on 
the feeder is not necessary, as the neu- 

lines where there are two or more 
phases present, since the difference of 
potential between phases is about 3800 
volts instead of 2200. This system re- 


FIG. 6. 

Feepek CE»rsj? 

: Th»C£ &HA3C *fAJH> 
~- SiMGLCPfrASE. Main* 

required for a three-wire three-phase 
system at 2200 volts under equivalent 

Standard 2200-volt transformers 
may be used for all purposes, being Y 
connected for power purposes and fed 
from a phase wire and neutral for 
lighting purposes. 

6. Combination systems. 

Such systems consist in the use of 
two or more of the above-described 
systems in combination. In direct- 
current systems the combination is 
usually made with a three-phase alter- 
nating system in order to facilitate 
transmission of electricity in large 
quantities. In alternating systems 
combination is sometimes necessary 
between two systems operating at dif- 
ferent frequencies, or with a direct- 
current system. 

Combination systems are usually the 
result of the development of the local 
distribution to a point where it be- 
comes more economical to feed the 
mains from two or more sources of 
supply. Sufficient is saved in the re- 
duction of the length and size of the 
distributing feeders to provide capital 
with which to erect substations and in- 
stall transmission lines. The conver- 
sion losses are not excessive, and the 
ability to concentrate all the genera- 
ting plants into one results in much 
better economy of production. 

tral wire carries the out-of-balance 
current, and the drop may easily be 
compensated for by the use of line- 
drop compensators in one of the four 
wires of the feeder. 

Installations of 5 to 20 h.p. may be 
supplied from two phases and the neu- 
tral by the open-delta transformer 
connection without interfering with 
regulation, as in the three-wire three- 
phase system. 

The arrangement of feeder and 
mains in this system is shown in 

Fig- /• 

The especial advantage of this sys- 
tem is. that when there is sufficient load 
to require a three-phase feeder the 
transmission is effected at 3800 volts, 
and loads up to 500 kw. may be dis- 
tributed from a single feeder at dis- 
tances of over three miles from the 
station with four wires, whose size is 
fixed only by their current carrying 

The neutral wire in this system 
naturally runs near earth potential and 
is therefore usually grounded at the 
generating station. This makes it 
necessary to look after the insulation 
of lighting-arrester cases, cables at 
points where they join overhead 
wires, fuse boxes and other fittings 
somewhat more carefully than in other 
systems. It is also necessary that line- 
men exercise more care in working on 





FctLC-C m Cs. A/ TEJt 

ZfouMk/mK lhA*M.E Phase. ^excatajs 

\ '*C^£ /?***£. fJE-ZMZJf 

\ Fovtr ¥$***- 7Z**C£. Pa*aS£- A/a»a/ 

fig. 7. 

quires one-third the copper required Both motor generators and syn- 

for a single- or two-phase system at chronous converters, commonly called 
2200 volts, or 44.4 per cent, of that rotary converters, are used in the con- 



February, 1908 

version of the alternating current to 
direct current. 

Where systems operating at two fre- 
quencies must be combined a motor 
generator is required, the motor being 
either of the synchronous or induction 
type. The synchronous motor may be 
wound for the transmission voltage 
and therefore requires no step-down 
transformer unless the transmission 
voltage exceeds 13,000. 

The commonest form of combina- 
tion system is that which generates 
alternating current for transmission 
and converts it to direct current at 
substations, by means of synchronous 
converters, for purposes of distribu- 

The superior economy of the syn- 
chronous converter outweighs its dis- 
advantages in most localities, and the 
large direct-current systems in cities 
like New York and Chicago utilize 
synchronous converters in preference 
to motor generators. 

The transmission systems in such 
cities are therefore operated at 25 
cycles in order to insure freedom from 
trouble with hunting or flashing of 
rotary converters. 

The portion of the load which is 
distributed as alternating current must 
then be generated by separate 60- 
cycle steam-driven generators, or by 
25-cycle motor-driven generators, 
since 25-cycle current is not well 
adapted to lighting, and is therefore 
not as salable as 60-cycle electricity. 

Where the 60-cycle distribution pre- 
dominates electricity is preferably gen- 
erated at this frequency, and the por- 
tion required in the form of direct 
current is secured through motor gen- 

The selection of the transmission 
frequency is therefore usually gov- 
erned by the relative amount of direct 
and alternating-current load, which is 
to be distributed. 

In cities like New York and 
Chicago, where over 75 per cent, of 
the distribution of electricity is 
effected by the use of direct current, 
the larger portion of the station ca- 
pacity generates 25-cycle current, the 
60-cycle current for outlying districts 
being derived chiefly from motor- 
driven 60-cycle generators. 

The voltage selected for the trans- 
mission from generating station to 
substations should be such that the 
cross-section of cables, bus bars and 
oil switches will not be excessive. The 
distances are usually so short in such 
transmissions that the cables mav be 
loaded up to their safe current carry- 
ing capacity without exceeding a con- 
servative percentage of line loss. 

As substation loads increase it is 
therefore desirable to have the volt- 
age high enough so that the saving 
in the cost of conductors and switches 

will more than offset the cost of extra 
insulation and safeguards incident to 
the use of the higher voltages. 

In the development of such systems 
in American cities, voltages ranging 
from 6600 to 13,200 have been 
adopted. The lower voltages were 
adopted during periods when the state 
of the art of generator design was 
such that higher voltages were not 
permissible, though desirable from the 
standpoint of economy in the cost of 
cables, etc. 

The limit of generator voltage has 
gradually been increasing, until at the 
present time manufacturers are pre- 
pared to wind generators for 20,000 
volts, in units of 1500 kw. and upward. 

Fig. 8 illustrates a typical distribu- 
tion system of this class. A low-ten- 
sion net-work is supplied from three 
substations by feeders, which are indi- 

been one of the controlling factors in 
the extension of the direct-current sys- 
tems of the companies operating in the 
larger cities. 

The battery reserve is invaluable in 
a city where many thousands of people 
are dependent upon continuous electric 

7. Series distribution. 

In this system all lamps are con- 
nected in series and operated at con- 
stant current, the voltage being varied 
automatically to maintain the current 
constant as the load varies. 

This system has been quite generally 
used for street lighting where it is 
economical of copper as compared 
with ordinary multiple systems on ac- 
count of the scattered nature of such 

Where open arc lamps are used the 

fig. 8. 

cated diagrammatically as radiating to 
intersecting points. The supply of di- 
rect current is derived from high po- 
tential lines through suitable convert- 
ing apparatus in each substation. 
These transmission lines are so ar- 
ranged that each substation has two 
sources of supply via different routes, 
thus insuring the substation against 
extended interruption of service due 
to failure of a transmission cable. 

The substation nearest the center 
of the net-work may be provided with 
a storage battery auxiliary as a re- 
serve in case of failure of any link in 
the chain of supply. It is also very 
useful as of assistance in obtaining 
proper pressure regulation and as an 
auxiliary source of electricity during 
the hour of the maximum load of the 

This feature of the direct-current 
system of supply is not easily appli- 
cable to alternating systems and has 

current strength is usually 7 or 10 
amperes, the lamps consuming 50 volts 
at their terminals, circuits being loaded 
up to not more than 125 to 150 lamps. 

With enclosed lamps the current 
strength is from 4.5 to 7 amperes, and 
the arc voltage about 70 volts. The 
voltage being limited by considerations 
of insulation in the lamps and circuit 
fittings, the maximum number of en- 
closed arc lamps on a circuit can be 
only 70 per cent, as many as can be 
carried with open lamps. The earlier 
installations were equipped with open 
lamps, but upon the development of 
the enclosed arc lamp, with its saving 
in cost of trimming. American practice 
has turned strongly to enclosed lamps, 
in spite of the reduction in circuit ca- 
pacity thereby involved. 

In Europe the lesser cost of trim- 
ming and the superior illuminating 
efficiency of the open arc have resulted 
in the retention of the open arc. 

February, 1908 



The high voltage makes this sys- 
tem unfit for general distribution pur- 
poses, and the constant potential sys- 
tems have therefore been universally 
employed for general distribution pur- 

The economy of copper inherent to 
the series system is greatly reduced 
where the lighting is not scattered over 
a wide area. The large number of 
circuits required for a heavy load in a 
small area requires a large investment 
in ducts, cables and pole space, and 
soon reaches a point where the value 
of pole space and ducts occupied by 
the extra circuits offsets the iarger 
section of copper in the single circuit 
of the multiple system. 

The earlier series systems, which 
reached commercial form before the 
incandescent lamps, were designed 
to be operated on direct current pro- 
duced by a special generator, which 
supplied a single circuit, or at most 
two circuits. The capacity of these 
generators was limited by the voltage, 
which, increasing with each additional 
lamp, prevented more than 125 lamps 
being carried on a single generator. 

With the development of alternat- 
ing-current systems, the expensive op- 
eration of a large number of arc 

generators called attention to the pos- 
sibility of the use of alternating cur- 
rent for series circuits. 

Two general types of equipment 
were developed for this purpose. In 
one of these a transformer, taking the 
current at the bus voltage, is provided 
with movable secondary windings, 
which are so counter weighted as to 
automatically vary the voltage im- 
pressed on the circuit, and thus main- 
tain the current constant. 

In the other type a constant po- 
tential transformer is used, but the 
voltage impressed on the lamp circuit 
is varied by a choke coil in the circuit. 
so designed that a system of weights 
will keep the current constant. 

Each circuit requires a transformer 
in the first type. A choke coil for each 
circuit and sufficient transformer ca- 
pacity for all or a group of circuits 
in one unit is required for the latter 
type. The latter type is somewhat 
less expensive, more efficient and has a 
higher power-factor than the former, 
and is therefore in more general use. 

In the alternating-current arc sys- 
tem the electricity may be generated 
by large and efficient units, thus avoid- 
ing the use of belts and shafting and 
an inefficient type of generator. 

Recently a new type of lamp using 
a metal upper electrode and a "carbon" 
of "magnetite" has been developed 
This lamp burns with a flaming arc 
of very high efficiency, and though not 
enclosed burns as long as the enclosed 
carbon lamp at one trimming. It gives 
about the same illumination with an 
expenditure of 320 watts that is given 
by a 10 ampere 500 watt carbon arc. 
It is essentially a direct-current lamp, 
however, and therefore requires some 
form of converting apparatus, such as 
a mercury arc rectifier, if it is used 
with alternating-current supply. 

Series systems using incandescent 
lamps are used for street lighting and 
in special cases where no other serv- 
ice is available in buildings. Such 
systems are of value in such places as 
residence streets when foliage inter- 
feres with effective distribution by 
means of arc lamps and where a 
smaller unit of illumination is ade- 

The advent of more efficient types of 
incandescent lamp, such as the tungs- 
ten, wilj doubtless result in the use of 
series incandescent systems much 
more generally than heretofore. 

Electric Locomotive Rail Pressure 

Experience indicates that the oper- 
ation of electric locomotives, owing to 
their lower center of gravity, has an 
effect upon the track entirely different 
from that due to the action of steam 
engines. In order to ascertain the ex- 
act nature and extent of this pressure 
upon the rails, the motive power de- 
partment of the Pennsylvania Rail- 
road has devised the apparatus which 
is being utilized at Clayton. 

A stretch of track about 166 ft. in 
length has been equipped with rails 
and cast steel ties, designed and made 
especially for this purpose. Instead 
of attaching the rail to the ties by 
spikes, a special form of block has 
been substituted, which allows a 
slight movement of the rail as the en- 
gine goes over it ; this movement 
registers the force with which the 
flanges of the wheels strike or press 
against the rails. It is expected that 
a large number of experiments with 
this apparatus will show the company 
quite accurately what the effect is of 
either steam or electric locomotives, 
moving at different speeds over either 
straight or curved track. 

Necessarily to make these tests, the 
engines must move at different speeds, 

and at all times each attains its maxi- 
mum speed. 

An electric apparatus has been de- 
vised to measure the precise amount 
of time elapsing while the different lo- 
comotives pass over this 166 ft. of 
track, in order that in computing the 
effect upon the track the exact speed 
attained may be known. The steam 
and electric locomotives, however, go 
over the track at different times, and 
there is no element of contest as to 
speed between the two types. The 
matter of speed is purely incidental to 
the main purpose of the tests, which is 
to enable the company, in planning its 
electric installations in New York, to 
design a track so safe as to be abso- 
lutely secure against any form of lo- 
comotive that mav be utilized. 

French Underground Cables 

that French cities require under- 
Skinner calls attention to the fact 
In a report Consul-General R. P. 
ground methods of distribution of 
electrical energy and that the new 
Marseilles cables are being laid in 
trenches like so much gas-pipe and 
with less expense and trouble, To 

satisfy inquiries in regard to this sub- 
ject he adds : 

These modern cables, although 
pliant and easily handled, are really 
impermeable conduits of small diam- 
eter. All that I have seen are manu- 
factured at Belfort and are delivered 
on huge wooden spools, from which 
they are unwound into the shallow 
trenches made ready for them with 
surprising rapidity. The copper wires 
composing these cables, arranged in 
groups of three, are first wound with 
jute, and the proper number of these 
groups is then wound again with jute, 
this cable passing next through an in- 
sulating bath. From this bath the 
cable passes through a lead press, 
from which it issues completely cov- 
ered with a thin lead sheathing. The 
lead sheath is now covered with jute, 
and the cable then enters a coal-tar 
bath, passing next through a final bath 
of lime, after which it is wound upon 
the wooden bobbin, the lime prevent- 
ing the tarry cable from adhering. 
The copper wires are presumed to be 
as secure from injury and deteriora- 
tion in their lead sheath as they would 
be in a costly tunnel or permanent 
metal conduit. 

A Peculiar Turbine Trouble 


MOST peculiar turbine accident 
lately occurred at the power sta- 
tion of the Los Angeles Gas & 
Electric Company, Los Angeles, Cal. 
It brought out some interesting" 
engineering points. As nearly as can 
be ascertained, two 2000-kw. vertical 
type turbines were operating in paral- 

of the machine. The steam valves 
were open full. Almost at once buck- 
ets stripped from their discs and 
broke through the shell of the tur- 
bine. The diaphragm buckled heavily 
and the flying debris cut the steam 
and oil piping and disabled the pumps. 
As a result of the breakage of the oil 


under the floor of the station with the 
generator end of the outfit project- 
ing above into the station room. 
There might otherwise have been 
considerable damage to station appa- 
ratus, not to mention the serious con- 
sequences to station operators present 
in the room above. 

It is well known to engineers fa- 
miliar with this type of machine that 
the peripheral veli tcity of the moving- 
outermost parts reaches in some ma- 
chines a velocity of nearly six miles 
per minute. At this speed the parts 
are under a very high tensile stress, 
which approaches 25.000 lbs. per sq. 
in. As the tension increases with the 
square of the velocity, it does not re- 
quire a very great increase above this 
speed to reach the elastic limit of 
steel. In fact, an increase of about 
50 per cent, in speed will raise the 
stress to 00.000 lbs. per sq. in., which 
is the ordinary breaking point of cast 
steel. It is probable that this point 
was reached after the failure of the 

lei. There had been considerable 
surging of the load back and forth 
between the two machines, due, it is 
believed, to the sticking of the valves 
of the governor. Suddenly one of 
the machines let go, throwing the 
governor from its position at the top 


pump, the pressure of oil at the step 
bearing of the second machine was 
lost and it came down on its step and 
was out of action also, though not 
otherwise essentially damaged. 

Fortunately, at the time, nobody 
was in the turbine room, which is 

The hunting of governors is not 
unknown even in steam-engine prac- 
tice, but it occurs in this type of ma- 
chine mainly under conditions of light 
load. The governor of the vertical 
type of turbine consists of an as- 
semblage of admission valves tripped 
by cams so arranged that their crests 
form a spiral contour. The valves 
open in succession and are. of course, 
cither entirely closed or entirely open. 
Thus, if there be six primary valves 
all tripping at full load, each valve is 
contributing one-sixth of the total en- 
ergy. If the load be running light 
two valves may be admitting all tin- 
steam, dividing the load between 
them. Even one valve may be con- 
trolling steam for the entire machine. 
When there comes an increase of load, 
another valve trips into action, and 
another one follows and they drop off 
as the load dies oft. For example, 
in a 2400-kw. turbine with 12 valves, 
each valve would admit steam equiva- 
lent to 200 kw. Suppose that the 
load on the machine is actually 1800 
kw. at a particular moment; nine 
valves will be admitting steam. If 
the load increases to 1900, the tenth 
valve will begin to admit steam and 
the rotating spindle will now get the 
impact of 200 additional kilowatts 
steam energy, which is 100 kw. more 
than the load requires. We have no 
longer a kinetic equilibrium between 
the working steam and the dragging 
load. We now have a uniformly ro- 

February, 1908 



fating mass subjected to an accelera- 
ting- force equivalent to ioo kw. at the 
speed of the machine. This relatively 
large force produces an actual in- 
crease in speed and stores up energy 
in the moving turbine spindle. If 
now the load drops suddenly the ki- 
netic energy of the moving spindle 
is sufficient to carry the load for a 
while, and one or more of the valves 
will cease admitting steam. The ma- 
chine will then work temporarily un- 
der its load while the moving spindle 
is giving up a part of its stored en- 
ergy. The governor will not readmit 
steam until the speed has fallen some- 

Suppose now there is a second ma- 

chine running in parallel with the 
first one. When the first one begins 
to drop its load it will be picked up 
by the second machine and the trans- 
fer of energy will continue until the 
governor of the first machine checks 
up the transfer of energy, when the 
process will begin to reverse. Xow 
all of this does not usually take place, 
except on a miniature scale, but the 
movements of energy may enormous- 
ly increase if, for any reason, the 
valves are sluggish in action or actu- 
ally stick. 

Owing to the fact that valve ad- 
missions of steam energy are by in- 
tegral parts of a considerable amount 
(200 kw. in the above illustration), 

there is undoubtedly a hunting back 
and forth by the governor for the 
proper nozzle area and the amount 
of hunting is not inconsiderable for a 
variable load. 

The rate of the governor move- 
ment is fixed, among other things, by 
what may be called the personal equa- 
tion of the machine. There is a defi- 
nite rate at which it can progress, and 
if a surging of the load falls approxi- 
mately into this periodicity and 
swings with it, we have a possible 
condition which would seem to have 
actually occurred in this case, where 
a surging of machines not overloaded 
preceded the trouble. 


AT the last meeting of the Insti- 
tute the committee on electrol- 
ysis, for which I had the privi- 
lege of acting as consulting electrical 
engineer, presented a report which 
contained abstracts of the most im- 
portant reports and papers on the 
subject of electrolysis from stray cur- 
rents published in America and 
Europe, together with conclusions 
and recommendations for the pro- 
tection of underground pipes, and 
also a brief outline of the theory of 
electrolysis. The reports of electrol- 
ysis tests which have been published 
since then show generally the same 
results as the ones which were quoted. 
I can endorse the conclusions and 
recommendations reached by the 
Committee as strongly if not more 
strongly to-day than when they were 

In the discussion of the theory of 
electrolysis it was stated that wher- 
ever current passes from a pipe to 
ground electrolytic corrosion takes 
place, and that a mere fraction of a 
volt may produce corrosion. It has 
been claimed by others that a voltage 
below \y 2 volts cannot produce cor- 
rosion because it is below the dissocia- 
tion voltage of water ; this, however. 
is a mistake, because when current 
passes from iron through ground to 
iron and produces corrosion the 
counter electromotive force is only a 
small fraction of a volt. If current 
should leave an iron pipe in ground 
containing certain alkalies the iron 

Xote. — A lecture delivered before the American 
Gas Institute. Washington, D. C, October 17, 


would be in what is known in chem- 
istry as the "passive'' state, and oxy- 
gen would be liberated without pro- 
ducing corrosion. In this case there 
would be a counter electromotive 
force of i l / 2 volts, due to the dissocia- 
tion of water, opposing the flow of 
current in such passive places. 

This passive condition of iron has 
also been put forward as a source of 
protection for iron pipes. The ques- 
tion has been investigated by Pro- 
fessor Haber,* of Karlsruhe, who has 
found that this passive condition of 
iron rarely exists in the case of iron 
pipes in ground. The electrolytic 
effect of current leaving passive iron 
would in fact be to change the alka- 
line to an acid condition and so render 
the iron active, that is, subject to cor- 
rosion. If there should be both active 
and passive portions in a positive pipe 
the current would leave largely in the 
active places because in the passive 
places there is the opposing electro- 
motive force of 1 3/2 volts, so that a 
corresponding increased densitv of 
current and of corrosion is set up in 
the active places where there is prac- 
tically no counter electromotive force. 
I cannot see, therefore, that we can 
derive any encouragement for the 
protection of pipes from this theory 
of the passive state of iron. 

A number of laboratorv investiga- 
tions on the effect of alternating cur- 
rent in producing electrolysis have 
also been published, and these indicate 
that an alternating current will pro- 
duce electrolytic corrosion which is 
very much dependent upon the com- 

*See Zcitschrift Fur Elcktrochemic, January 26, 

position of the soil, and which varies 
in amount from a fraction of one per 
cent, to possibly two per cent, of the 
corrosion produced by an equivalent 
direct current. It would appear from 
this that in certain cases, particularly 
in the case of a lead pipe, even an al- 
ternating current may produce de- 
structive corrosion. No experiences 
with stray alternating currents have, 
however, as yet been made public. 

In Germany a special form of elec- 
trode for measuring earth potentials, 
which does not introduce uncertain 
polarization potentials, has been used, 
and also a special apparatus for 
directly measuring the current flow- 
ing through ground. These were de- 
vised by Professor Haber, who has 
called them "non-polarizable elec- 
trode" and "earth amperemeter" re- 
spectively : he has used them in an ex- 
tensive series of tests upon the piping 
system of Karlsruhe. A description 
of the electrode and earth ampere- 
meter, together with the results of 
these tests, were presented in a paper 
before the Deutscher Verein von Gas 
und Wasserfachmannern, which is 
published in their transactions for 
1906. These devices have also been 
used with success during the past year 
in the tests made by the German 
electrolysis committee. 

The mode of testing a piping sys- 
tem for stray currents is of consider- 
able importance and interest to gas 
engineers. A- this was not touched 
upon in the report of last year. I will 
in what follows endeavor to give an 
outline of suitable methods for ma- 
king such tests, and also briefly de- 
scribe the construction and uses of 



February, 1908 

the new German devices. I will lay 
special stress on methods which will 
serve to prove that corrosion was due 
to stray currents and to prove the 
origin of these currents so as to fix 
the responsibility, as this is important 
and essential evidence in case of a law- 
suit ! 

With the arrangement, usual with 
trolley roads, of connecting the neg- 
ative terminal of the generator to the 
rails, the general path of the stray 
currents is from the rails through 
ground to the pipes at places distant 
from the power-house, through the 
pipes, and from these through the 
ground back to the rails or to other 
return conductors in the vicinity of 
the power-house. Where current 
flows from a rail to a buried pipe the 
rail assumes a positive potential with 
reference to the pipe ; where current 
flows from a buried pipe to a rail the 
pipe assumes a positive potential with 
reference to the rail. 

The first step in making an electrol- 
ysis survey of a piping system is 
therefore to measure potential differ- 
ences between pipes and rails in a 
large number of places throughout 
the system in order to locate these 
points of current flow between pipes 
and rails. As gas mains are not gen- 
erally accessible, service or drip con- 
nections are used for making connec- 
tion with the voltmeter; for this pur- 
pose one voltmeter wire is clamped or 
otherwise fastened to the drip or 
service connection, care being taken 
to clean with a file the part where the 
wire is fastened so as to insure good 
metallic contact. The other voltmeter 
wire is best soldered to a rough flat 
file which is held on the rail for the 
contact, and which can be quickly 
removed and replaced when a car 
passes. The voltmeter for these 
measurements should have a high 
resistance so that an accidental poor 
contact at a drip or service connec- 
tion will not seriously interfere with 
the measurement. A suitable instru- 
ment is a portable high resistance 
Weston voltmeter with zero center, 
having ranges of 1.5, 15 and 150 
volts. Readings should be taken 
every 10 seconds for 10 minutes at 
each point, and the maximum, mini- 
mum and average reading noted. 

The average readings are then con- 
veniently marked upon a map show- 
ing the principal pipes and tracks, red 
numbers being used where the pipes 
are positive to the rails and black or 
blue numbers where these are neg- 
ative. An excellent plan is also to 
plot these potential differences graph- 
ically upon a map on which the pipes 
are shown as lines and using these 
lines for axes, and the voltmeter read- 
ings for ordinates; by shading the 
areas between the potential curves 

and the pipe lines with red where the 
pipe is positive and with black or blue 
where it is negative to the rails, a 
clear representation of the potential 
distribution is obtained. If the neg- 
ative bus bar of the power station is 
connected to ground plates or to other 
buried metal as cable sheaths, meas- 
urements of potential differences be- 
tween these and the pipes should also 
be made and plotted upon a separate 

It will generally be found that there 
are definite regions in the neighbor- 
hood of supply stations and of return 
feeder connections in which the pipes 
are always positive to the rails ; and 
that there are other definite regions, 
remote from the first, in which the 
pipes are always negative to the rails. 
Between the positive and negative 
regions the potentials will be found to 
fluctuate from positive to negative. 

The existence of potential differ- 
ences between pipes and rails, even if 
large, is however no conclusive evi- 
dence of stray currents, but indicates 
at what points current may be flowing 
from rails to pipes and at what points 
it may be flowing from pipes to rails 
or to other return conductors. A 

covered wires so as to have a good 
insulation and to prevent the wire, if 
coming in contact with wet ground, 
from taking the potential of the 
ground and disturbing the readings. 
A good wire for this is No. 14 rubber- 
covered and double-braided wire. I 
remember one instance where we were 
measuring the drop along a cable 
sheath on a wet day, getting readings 
of from one to two millivolts ; during 
one of these readings the needle sud- 
denly ran off the scale, and a volt- 
meter which was substituted indicated 
two volts. We found upon investiga- 
tion the connecting wire lying in 
water, the insulation defective, and 
that we were therefore getting the 
potential difference between the cable 
sheath and the surface of the ground 
instead of the drop in the cable sheath. 
A study of the map will show at 
what points determinations of current 
strength should be made. This is 
done by measuring the drop in poten- 
tial between two points in the pipe by 
means of a millivoltmeter, and divid- 
ing this by the resistance of the in- 
cluded length of pipe. If this length 
contains one or more joints this re- 
sistance must be measured and not 

FIG. I. 


high potential difference is, in fact, 
usually a sign that there is a high 
ground resistance and but little cur- 
rent flowing. 

The next step is to determine the 
direction of the probable current 
flowing in the pipes. This is done by 
measuring potential differences be- 
tween two points in a pipe by means 
of a millivoltmeter. A convenient in- 
strument for these measurements is a 
zero center Weston millivoltmeter 
with two scales, one of 10 and the 
other of 100 millivolts. These meas- 
urements may be made between drips 
or service connections from 50 to 200 
feet apart. These measurements can- 
not be used, however, for calculating 
the current strength in the pipes but 
only to indicate the probable existence 
and direction of this current. This 
direction of flow is then marked upon 
the map together with the potential 

In making these tests it is im- 
portant that the connecting leads used 
with the millivoltmeter be rubber- 

estimated, because joints make con- 
tacts of extremely variable resistance. 
As this resistance measurement is 
troublesome to make, it is generally 
more convenient to measure the drop 
in potential between two points in one 
continuous length of pipe. A sketch 
of this arrangement is shown in Fig. 
1. With very large pipes and small 
current this drop is a fraction of a 
millivolt, and requires a specially 
sensitive millivoltmeter reading to 
hundreds of a millivolt for its meas- 
urement. The resistance of the length 
of pipe can be calculated from its 
dimensions and from an assumed fig- 
ure for the conductivity. Mr. Maury, 
in a paper* before the American 
Water Works Association, has given 
convenient tables for converting drop 
in millivolts directly into amperes of 
current for cast iron and for wrought 
iron pipes. These tables are based 
upon a resistance of 0.00144 ohm per 

'"Surveys for Electrolysis and Their Results." 
A paper read at the Twenty-third Annual Con- 
vention of the American Water Works Associa- 
tion, held at Detroit, Mich., June 23-26, 1903. 

February, 1908 



pound-foot for cast iron and of 
0.000181 ohm per pound-foot for 
wrought iron ; these figures were ob- 
tained from measurements upon a 
large number of pipes, and agree well 
with these given by others and with 
measurements made by the writer. 
To find the resistance of any cast or 
wrought iron pipe per foot of length, 
it is only necessary to divide the above 
figures by the weight of the pipe per 
foot. In the accompanying table I 
have given the weights and resist- 
ances per foot for the usual sizes of 
cast and wrought-iron pipes, the re- 
sistances being based upon the above 
figures for the pound-foot. 

tained by soldering the connecting 
wire directly to the pipe ; this is par- 
ticularly advantageous when readings 
are to be taken over a considerable 
time. This soldering should be done 
by means of a torch and sufficient 
time allowed for the joint to cool to 
the temperature of the pipe before 
using it. as otherwise a thermo-electro- 
motive force may be set up which dis- 
turbs the reading. When such con- 
tact wires have been soldered to a 
pipe it is convenient, whenever prac- 
ticable, to continue these with rubber- 
covered wires to the surface, or to 
some accessible point, and to ter- 
minate the ends in small iron boxes 



Cast Iron 

Wrought Iron 

Extra Heavy 
Wrought Iron 

of Pipe 


per Foot 

without Hub 



per Foot 



per Foot 

without Hub 



per Foot 


per Foot 
without Hub 

per Foot 

l A 

.84 J 


2.7 « 


7.5 > 
10.6 if 
18.8 _ 
28. 1 
40. i 











2 2 

10. i 




















11 . 































A matter of great importance in 
making these current measurements 
is to be sure to have perfect metallic 
contact between the pipe and the 
millivoltmeter wires. It will not do 
to use drip or other fittings for this 
connection, but contact must be made 
directly with the metal of the pipe. 
It is therefore necessary to expose the 
pipe where a current measurement is 
to be made. 

I have made tests with various 
methods of making contacts for cur- 
rent measurements upon a rusty pipe, 
and have found that the presence of 
oxid may produce such a high resist- 
ance as to prevent a reading. I have 
also found that moisture can produce 
an electromotive force due to electro- 
chemical action so large as to entirely 
offset the reading due to drop by the 
current. A satisfactory method is to 
use a pointed piece of steel about half 
the size of an ordinary lead pencil, 
with the connecting wire soldered to 
it, and provided with a wooden 
handle, the soldered joint being inside 
of the handle ; this steel contact is 
pressed against a spot on the pipe 
which has been previously filed bright. 
Bv far the best contact is however ob- 

such as drip boxes ; these wires are 
then available for future current 
measurements on the pipe without 
going to the labor of again exposing 

These readings of current should 
be taken every 10 seconds for at 
least 10 minutes, and the maximum, 
minimum and average readings noted. 

By tracing the flow of currents 
found in the pipes from these meas- 
urements points can be usually located 
at which current must be leaving the 
pipes, and the pipes should be exposed 
here and examined for evidences of 
electrolytic corrosion. It must be 
remembered that all current which is 
found flowing in a pipe must leave it 
somewhere in order to return to the 
negative pole of the generator. This 
follows from the fact that every elec- 
tric circuit must be completely closed 
so that every ampere which leaves the 
positive pole of the generator must 
eventually return to the negative pole, 
no matter how long or complicated 
the path through which it passes may 
be. Electricity is in this respect very 
different from gas or water, which 
latter may escape from a leak and dif- 
fuse through the ground. It is for 

this reason that stray currents repre- 
sent no loss of electricity to the rail- 
road, as is sometimes supposed, but 
are, on the contrary, a gain, inasmuch 
a- tlie ground and pipes offer addi- 
tional return paths for the railway 
current and therefore mean less drop 
in the return circuit. It is for the 
purpose of taking advantage of the 
earth as a return conductor that rail- 
way power stations frequently use 
large ground plates connected to the 
negative bus bar with heavy copper 

The previous brief outline of tests 
will serve to locate the points in the 
positive district in which there is the 
greatest danger from electrolytic cor- 
rosion, and these tests are relatively 
easily made and constitute the elec- 
trolysis survey as ordinarily made in 

There may, however, be many other 
endangered points which these simple 
tests will not reveal. Stray currents 
do not always take the simple path 
from rail to pipe, along the pipe and 
back to rail or other return conductor, 
but frequently take roundabout paths, 
passing from one piping system to a 
second system, from this, perhaps, to 
a third system, or passing across 
pipes, shunting around high resist- 
ance joints, etc., producing electro- 
lytic corrosion at every point of leav- 
ing the pipe for ground. These 
points may exist in the negative, 
neutral or positive districts, and are 
much more difficult to locate by means 
of electrical measurements. 

The writer has in mind a striking 
case of this kind which he saw while 
witnessing some tests made by Mr. A. 
A. Knudson. in Rutherford, N. J. A 
water pipe crossing under trolley 
tracks was badly corroded, with all 
evidence that the corrosion was elec- 
trolytic, yet this pipe was always over 
one volt negative to the rails and 
therefore could not be giving up cur- 
rent to these rails. Upon further in- 
vestigation it was found that an oil 
pipe crossed this water pipe several 
feet away in the ground, and that the 
water pipe was highly positive to this 
oil pipe. It was therefore clear that 
the current which was producing the 
corrosion was leaving the water pipe 
and entering the oil pipe. 

Wherever there are two or more 
independent piping systems, as for in- 
stance water and gas pipes, measure- 
ments of potential difference between 
these should he made to see if any 
points can be located where current is 
likely to be passing from one to the 
other. Current measurements must 
then he made in the pipes at these 
points and plotted and studied as be- 

Measurements of current should 



February, 1908 

also be made at every point where a 
pipe crosses a trolley track or passes 
through particularly wet ground near 
trolley tracks, to see whether current 
is passing from one to the other, as 
these are particularly endangered 
points. For this purpose it is best 
to expose a length of pipe on each side 
of the track and to measure the cur- 
rent on each side. It may be found, 
however, that this current is so fluctu- 
ating that it is difficult to draw con- 
clusions from these readings ; the 

where it passes under the track, and 
therefore if current is passing from 
pipe to rail or from rail to pipe at 
this point. The same method may be 
applied in any case where it is desired 
to find out whether there is a change 
in the current flowing in a pipe be- 
tween any two points in the pipe. The 
illustration (Fig. 2) shows two such 
simultaneous current readings taken 
in a pipe on each side of a stretch of 
wet ground in which the pipe was 
laid for a distance of about 500 feet 

B, the curves showing that during 
most of the time the test current 
was entering the pipe from the wet 
ground between these stations. 

In making these simultaneous cur- 
rent measurements it is very impor- 
tant that the two millivoltmeters have 
the same period of oscillation, as 
otherwise, with the usual fluctuating 
current, one would lag behind the 
other in its reading, and the instanta- 
neous readings would no longer be 
comparable. It is desirable to use 

— - - 

— . _ 



in f^ 


at St 

a ti or 





Cu rr 

en T 1 

H /M 

ftrr 4 

f St, 
















1 \ 









1 i 



1 \ 

• \ 






; J 
• •• 






, 1 










1 T 














« 6 


\ 1 
1 1 

\ 1 

\ / 












\ / 

> 1 




in Af 

1 n vtt 


. Mil , d 

— £ 














-. 10 




passing of a car may produce an en- 
tire change in the current conditions. 
I have found that the best way to 
make the test under these circum- 
stances is to use one millivoltmeter 
on the pipe on each side of the track, 
and to take simultaneous readings of 
current every 10 seconds for 10 or 20 
minutes. The time of taking these 
readings should include the passing 
of a car. When these readings are 
plotted it can be seen by comparing 
the current curves whether there is a 
change in the current in the pipe 

and only one-half block away from 
trolley tracks ; the agreement of the 
current curves shows, however, that 
current was neither passing to nor 
from the pipe in this wet ground dur- 
ing the time of these tests. 

A case where current was entering 
a pipe from wet ground is shown by 
the curves in Fig. 3. in which the 
simultaneous current readings are 
plotted, taken on each side of the wet 
ground through which this pipe was 
passing. The direction of the current 
Mow was from Station A to Station 

two exactly similar meters, and these 
should be strongly damped so as to 
move slowly enough to enable accu- 
rate readings to be taken even with 
a very fluctuating current. 

Another very important matter is 
the choice of time for making these 
measurements. The stray currents in 
the pipe will follow the load curves 
of the railway, and tests should be 
made at time of heaviest load as well 
as at times of light load. 

In a large railway system operating 
a number of stations, some of the sta- 

February, 1908 



tions may be shut down during 
periods of light load, and the current 
condition in the pipes be completely 
changed during- these times. If the 
stray currents in the pipes are the 
combined leakage current from two 
or more railway systems, having dif- 
ferent load curves, the current condi- 
tions in the pipes will also change 
during different times of the day, and 
perhaps also on different days. In 
cases of this kind a recording meter 
should he applied to a number of 

as shown for Monday and Tuesday in 
Fig. 4. The diagram for Sunday 
shows that the current in the pipe was 
very large, and flowing practically all 
day in the same direction ; this large- 
current is accounted for by the fact 
that the neighboring trolleys were 
carrying a large crowd of Sunday ex- 
cursionists. The following Sunday 
was a rainy day, and the record 
showed a much smaller current in the 
pipe for this day. I want to say that 
the change in direction of the current 

the probable destruction. One am- 
pere leaving an iron pipe for wet soil 
will remove 20 pounds of iron in one 
year. But the same ampere of cur- 
rent in its path through the ground 
and pipes may leave the pipe and re- 
turn to it any number of times, shunt- 
ing, for instance, around occasional 
high-resistance joints, and removing 
iron at the rate of 20 lb. per year at 
every point of leaving the pipe. The 
actual damage is also very much de- 
pendent upon whether the exit of cur- 




characteristic points in the 
system, arranged to record the cur- 
rent in the pipe. I have used a Wm. 
H. Bristol recording meter (a pyro- 
meter instrument without its thermo 
couple) to obtain 24-hour records of 
currents in pipes. The diagram in 
Fig. 4 shows the current in a gas pipe 
for four consecutive days, 
from four Bristol charts 1>\ 
ing the average current for 
tive one-hour periods and 
these averages. The records for 
other week days were about the same 



indicated in the week-day diagrams 
is an unusual condition, and probably 
due to changes in the operation of the 
power-houses producing the currents 
affecting the pipes. These current 
records, by their agreement with the 
load curves of the trolley road, are 
frequently also useful in serving as 
evidence that the currents in the pipes 
are trolley-road currents. As the 
electrolytic damage is proportional to 
the time during which the current 
acts, it is necessary to have the time 
curve of current in order to estimate 

rent is concentrated in a small area or 
is distributed over a large area. Ex- 
perience shows that the electrolytic 
corrosion ordinarily takes place in 
spots, producing pittings in the pipes, 
which indicates that the current must 
have the pipes in these spots and not 
from the entire surface; the pipes are 
in this way much more quickly de- 
stroyed than if the corrosion took 
place uniformly over the entire sur- 
face. It therefore is an extremely 
difficult matter to estimate the amount 
oi electrolytic damage and such esti- 



February, 1908 

mates will generally be too low for 
the above reasons. 

It is possible to trace the path of 
current through the ground by meas- 
uring potential differences between 
different points in the ground. An 
iron rod driven into the ground is 
often used for making contact for 
such measurements. This, however, 
is not reliable because the electro- 
motive force of polarization of the 
iron in wet ground is dependent upon 
the surface condition of the iron and 
upon the ingredients of the ground ; 
it is not constant, and not necessarilv 

against the part of the ground at 
which the potential is to be measured, 
thus establishing contact between the 
ground and the zinc sulphate solu- 
tion. There is only a negligible polar- 
ization voltage between the zinc- 
sulphate and ground. The zinc in 
this solution has a definite and con- 
stant polarization voltage with refer- 
ence to the solution. When two of 
these electrodes are used to measure 
the potential difference between two 
points in the ground, the two polar- 
ization voltages balance each other. 
With these electrodes a zero method, 

by in the ground would be 0.45 volt, 
the pipe being positive. If then a 
measurement shows the pipe say 1.1 
volt positive to the electrode it means 
that the pipe is 1.1 — o.45=-|-o.65 volt 
(positive) to the ground and proves 
that current is flowing from the pipe 
to the ground. If a measurement 
shows a pipe say 0.15 volt positive to 
the electrode, it means that this pipe 
is 0.15 — 0.45= — 0.3 volt (negative) 
to ground and proves that current is 
flowing from the ground into the pipe. 
Ground potentials of less than one- 
tenth volt cannot safelv be used as 





r r t 


> in 


?s / 






v £ 








B V 

e 3t. 


< n<? 

a *e 


c u 
+ Vt 

t Ct 6; 

/s . ' 

i tn 

e or c 
1 ica 

n e 
te t 



?0 uj<3 "& A 'or A 9 



- ^c 


.. a 

wa y 




























i 5 




ft ; 

1 * 
1 * 






1 *• 
/ ^ 






\ x 















m 3 

u m/T 

«/- L 






t ura 

ay, / 

U<f ' 



n cfa,t 

', M<* 

y 19 


7 Cf*.lf 




'ay . 

\— t 




the same for each of two rods in dif- 
ferent parts of the ground. Where 
small potential differences are meas- 
ured between two points in the 
ground by means of iron rods as elec- 
tiodes entirely wrong results may be 
obtained because of difference in the 
polarization voltages at the two elec- 

To overcome this difficulty the 
non-polarizable electrode" shown in 
5 was devised by Professor 
Haber. This consists of a glass tube 
about two inches in diameter and 
10 in. long, having a porous cup 
:emented to one end, and containing a 
saturated solution of zinc sulphate 
with a zinc rod dipping into the solu- 
tion. The porous cup is pressed 


such as a potentiometer with a sensi- 
tive galvanometer, should be used for 
measuring the potential differences, 
because the high resistance of the 
contact through the porous cup 
would produce a drop in voltage and 
reduce the -reading on a direct-read- 
ing voltmeter. This non-polarizable 
electrode is also used to measure the 
potential difference between a pipe or 
rail and the ground ; in this case the 
polarization voltage of the electrode 
must be taken into account, which is 
equal to 0.45 volt negative with refer- 
ence to the iron of a pipe or rail. This 
means that without current flowing 
between pipe and ground the poten- 
tial difference between an iron pipe in 
ground and this electrode placed near- 

evidence of current flow as they may 
be due to accidental causes. 

It may be desirable to measure 
directly the flow of current through 
ground as between a pipe and a rail, 
or between two pipes. The best ar- 
rangement for this is the "earth am- 
peremeter" devised by Professor 
Haber and illustrated in Fig. 6. This 
consists of a wooden frame with two 
copper plates about four inches square 
and separated by a plate of mica or 
glass as shown in the figure. These 
copper plates are covered with a layer 
of a paste made of copper sulphate 
and water with 20 per cent, sulphuric 
acid, the thickness of paste being 
about 1 / 16 -inch. A wetted piece of 
parchment paper is laid over this 

February, 1908 



paste and the remainder of the frame 
is filled with soil taken from the por- 
tion of ground where the current is to 
be measured. This frame is then 
placed in a suitable excavation in the 
ground where current is to be meas- 
ured, so that the current flow is as 
nearly as possible normal to the 
frame, and the latter is then com- 
pletely covered up, the soil being 
tightly stamped in. Insulated copper 
wires are brought out from the plates 
and connected with a milliampere- 
meter ; this will indicate the current 
flowing through this section of the 
ground. The object of the copper 
sulphate paste is to prevent polariza- 
tion at the surfaces of the copper 
plates, which otherwise would pro- 
duce an electromotive force opposing 
the flow of current through the meter. 
The direction of the meter indication 
will give the direction of the current 
flow through the ground. It might 
lie supposed that the introduction of 
this frame would entirely change the 
current conditions in the ground ; this 
has not been found to be the case, due 
to the fact that the resistance of the 
path through the ground is verv higfh ; 
the displacement of the small portion 
of ground taken by the copper plates 
therefore does not greatly change the 
total resistance of the path of the cur- 
rent through the ground. The no- 
tion that the earth is a conductor of 
negligible resistance applies only 
when the path of the current is 
through an unlimited cross section ; 
the resistance of a limited section of 
earth is comparatively high, varying 
under ordinary conditions from IOO 
to iooo ohms between opposite faces 
of a cubic foot. 

This earth amperemeter is well 
suited for measuring current flow be- 
tween pipe and ground ; for this pur- 
pose the frame is buried in the ground 
from one to two inches from the pipe, 
and parallel to the pipe and normal 
to the probable current flow through 
ground. This measurement can be 
used to form an estimate of the prob- 
able amount of electrolytic damage to 

I want to call attention to a wholly 
incorrect method for determining cur 
rent flow between pipes and rails 

Copper W,r*. 

■ Copi//oruTu be 



— ■« — Glass Tube 

Porous Cub 


through ground, which I have seen 
described in a certain class of electrol- 
ysis reports. I refer to the method 
in which the potential difference be- 

tween a point in the rails and a point 
in the pipes is measured with a volt- 
meter, as between C and D in Fig. 7, 
and the resistance between these same 
points is measured with an ohmmeter; 
the voltage divided by the resistance 
1 given as the current in amperes 
flowing through the ground. This is 
then repeated for other points, as E 
and F, G and H, etc. (Fig. 7), and 
these currents are added up to give 
the total current through the ground. 
In one published report of this class 
upon a water piping system in a West- 
ern city, this total current is given as 
0148 amperes, with a calculated de- 
struction of 122,980 lb. of iron per 
year. The following words of warn- 
ing are found in this report: "The 
loss of 122,980 lb. of metal per year 
from the distributing water system 
must eventually bring great and seri- 
ous trouble, involving the expenditure 
of many thousand dollars." The ab- 
surdity of these figures is clear when 
we consider that many times this cur- 
rent could have been obtained by mak- 
ing tests in a sufficient number of ad- 
ditional points in the system. Cur- 
rent through ground between pipes 
and rails cannot be determined from 
such measurements of potential dif- 
ference and ground resistance. The 
latter is the joint resistance of all con- 
ducting paths between the entire pip- 
ing system and the entire rail system, 
while the potential difference is that 
between a point in the pipes and a 
point in the rails ; these two measure- 
ments can therefore not be combined 
by Ohm's law to give current, and the 
figures which have been given in the 
reports referred to are entirely with- 
out meaning; the results and con- 
clusions based upon them are conse- 
quently wholly untrue. 

As an example of the wholly in- 




1 ■ ' • ' ■■ "i ' 11 V ■ '• ■ - •■■ -"^ .'.•■>■•• > ■ ■'.-•.■ -.'. v^ - 


the pipe, and in cases where corrosion 
has taken place it will serve as evi- 
dence that it was caused by stray 

1 have made a number of laboratory 
tests with the non-polarizable elec- 
trode and the earth amperemeter and 
have obtained exceedingly satisfac- 
tory and consistent results. 







February, 1908 









D •+ 






February, 1908 



correct conclusion which this method 
may lead to, I have shown in Fig. 8 
the instantaneous values of current 
measured in a gas pipe at stations on 
each side of a stretch of ground 
soaked with salt water, taken simul- 
taneously with potential readings be- 
tween this pipe and trolley tracks close 
by and in contact with this wet 
ground. The gas pipe is seen to be 
about one volt (average) positive to 
the rail. The resistance between the 
rails and pipe must be very low so 
that a current calculated by the above 
method would give a large value and 
lead to the conclusion that this cur- 
rent was passing from the pipe to the 
wet ground and producing corrosion. 
The simultaneous current readings 
show, however, that there is prac- 
tically no escape of current from the 
pipe between these stations. This 
class of report is generally also full 
of remarkable statements which 
greatly exaggerate the electrolysis 
danger and which seem to be de- 
signed to scare people against the 
electrolysis evil. From the absurdity 
which appears on the face of these 
reports they tend to hide the real 
electrolytic conditions and the real 
dangers which exist. The technical 
press has already exposed and 
severely criticized some of these re- 
ports which have been published. 

Y\ 'here more than one electric road 
is operating in the vicinity of a pip- 
ing system, the stray current in the 
pipes may be the leakage current 
from one, or it may be the sum of the 
leakage currents from all of the roads. 
In cases where electrolytic damage 
has resulted it is important to trace 
and prove the source or sources of the 

destructive current in order that 
claims for damages and a demand to 
stop this current may be presented to 
the proper owners, and pressed in 
court if necessary. This is fairly 
simple if it is possible to disconnect 
one power-house at a time sufficiently 
long to measure the effect upon the 
pipe current. If this cannot be clone 
indirect methods must be used which 
may make the problem very trouble- 
some, especially in cases where two 
or more roads use a common track 
for part of their line. Noting the 
starting, running and stopping and 
location of the cars on each line, and 
comparing this with the current 
curves in the pipes, will often help to 
locate the source or sources of the 
pipe current. Where the tracks of 
the different roads are independent, 
a good method is to attempt to 
measure the current flowing through 
the ground between pipes and the 
rails of each system by means of the 
earth amperemeter. No universal 
method can, however, be given be- 
cause the conditions vary so widely ; 
it is necessary to study each case 
separately and to adapt the methods 
best suited to it. 

Of the attempted remedial schemes 
the one most frequently applied to 
pipes in America is bonding in the 
positive districts to the negative bus 
bar. While affording local protection, 
this greatly increases the amount of 
stray current, frequently producing 
electrolytic trouble in other portions 
of the piping system, as at joints, and 
the pipes themselves become a source 
of danger to other underground metal. 
In cases where there are variations in 
current flow, as was, for instance, 

illustrated in Fig. 4, no bonding 
scheme can be applied, as the positive 
zones keep shitting about. 

Insulating joints can only be used 
in special cases to effect a partial re- 
duction in the stray currents in the 
pipes, and often do more harm than 

I am convinced that in most cases 
there is but little, if anything, that 
the pipe owner can do to protect his 
pipes from these stray railway cur- 
rents. The cure must and in all fair- 
ness should come from the railways. 
This is the attitude which the Ger- 
mans have maintained ever since they 
have begun to study this problem. 

The question of the proper restric- 
tion to place upon the railroad in 
order to render underground piping 
reasonably safe from stray currents 
is a hard one to settle. In some places 
a maximum permissible potential dif- 
ference in the rails, or between rails 
and pipes, has been prescribed by city 
ordinance. The amount of stray cur- 
rent depends, however, just as much 
upon soil resistance as upon potential 
difference, and this resistance varies 
so largely that no maximum, safe 
voltage can be prescribed for all cases 
except one so low as to be entirely 
impracticable. The maximum allow- 
able current in any one pipe is some- 
times prescribed ; this, however, is 
also inadequate. The only really safe 
condition would be a maximum allow- 
able density of current leaving any 
one pipe for ground expressed in 
milliamperes per square foot of pipe 
surface. This is, however, trouble- 
some to measure, and the restriction 
would be very difficult to enforce. 

An Odd Case of a Motor Dropping Its Load 

A 10-h.p., no-volt, direct-current 
motor received its current from an 
engine-driven generator, the engine 
being also belted to drive shafting. 
The motor was connected to this line 
shafting by a chain drive. The pur- 
pose of this arrangement was to test 
power chains. 

By weakening the field of the motor 
its increased speed allowed it to return 
power to the shafting through its 
chain drive. The work done by the 
chain under test is varied by changing 
the field strength of the motor. 

After several years of successful 
operation in the above manner the 
motor became inoperative. Within a 
few minutes after starting it up and 
weakening its field to allow a flow of 
current corresponding to the torque 

required, it would be found that the 
current, for some unaccountable reason, 
had dropped nearly to zero and the 
motor would be simply floating on the 
line without transmitting any power 
to the shafting through the chain 
drive. There was about ^-inch end 
play of the shaft, and by pressing 
slightly on either end of the shaft so 
as to stop the end motion, the current 
taken by the motor and consequently 
its torque would raise immediately 
from zero to normal. The motor was 
then operated as a generator, receiv- 
ing its power from the shafting 
through the chain drive and feeding 
into a bank of lamps. With the shaft 
oscillating normally the voltage of the 
generator was no, but by prosing on 
the end of the shaft (thus preventing 

any end play) the voltage immediately 
rose to 115 volts. 

It was found that two of the four 
brush-holders were carrying double 
the amount of current carried by the 
other two, although a voltmeter placed 
across each of the four field coils 
showed the same voltage drop aero-.. 
each coil. 

The trouble was found in one of the 
field coils, which had a broken wire in 
its center. When the shaft was not 
allowed to oscillate the broken wire 
laid together and the motor operated 
normally, but when the shaft oscillated 
the vibration of the motor, due to the 
pounding of the collar on the motor 
shaft against the bearing, caused the 
perplexing action we have just no- 
ticed. A new field-coil cured this 

The Application of Electric Power to 
Pulp and Paper Mills 


THE wonderful progress made in 
the last decade, particularly dur- 
ing the last five years, in the 
application of electricity as a motive 
power to all branches of commerce 


in the business, and I am 
ashamed to admit it, but we are wa- 
king up, and in a few months' time you 

won't hud any conditions like this, in 
my mill at least. The trouble with us 


(1 the in- 

reason why we should change these 
conditions. We simply let well 
enough alone." 

The above statement undoubtedly 
oilers the onl_\- possible excuse for the 
existence of such poor devices for 
power transmission in an age when 
much better ones are realized facts, 
but with the rapid increase in the 
number of paper mills within the last 
few years, and the keener competition 
resulting therefrom, it has become a 
necessity to reduce the cost of pro- 
duction, and this without lowering 
the quality of the product. To secure 
low cost of production means not only 
an economical administration of all 
the affairs of the mill, but also the 
adoption of the modern labor and 
power-saving apparatus and devices. 
The best managed mill in the country, 
if it is handicapped by an antiquated 
equipment, cannot hope to compete 
with a modern mill operating under 
improved conditions. 

Xo line of improvement in paper- 
mill equipment will contribute more 


and industry, has aw; 
terest of the whole 
world to its possibilities, and the 
paper-mill operator, like his brother 
manufacturers in other lines, is to- 
day making a conscientious investi- 
gation into the results to be secured 
b\ the application of this power to his 
own particular branch of industry. 

The scheme of power transmission 
which has prevailed in paper mills 
lias a parallel in that t<> be met with 
in any other factory of equal or 
greater size not already equipped with 
electric motors. A distinguishing dif- 
ference, however, between the paper 
mill and the other types of factories 
is found in the fact that in possibly 
in other industrial establishment is 
the percentage of losses due to an in- 
efficient system of power transmis- 
sion so great as in the paper mill: 
and. as a consequence, in no other 
type of industrial establishment is the 

per cent, of 


resulting 1 from 

electrical equipment so large. A short 
time ago a prominent paper manu- 
facturer of the Middle West said, in 
answer to an expression of surprise 
from us on discovering some excep- 
tionally inefficient drives in the power 
transmission of his mill: "I don't 
wonder at your surprise. I know- 
that we paper men are the worst old 



has been that in the past we did so 
well, even under unfavorable condi- 
tions, that there was no apparent 

to the securing of low cost of produc- 
tion and high-grade output than the 
adoption of the electric system of 

February, 1908 



power transmission. This is a very 
broad statement, but it can be sub- 
stantiated by a careful investigation 
of the results obtained from mills 
which are actually equipped with the 
electric system. Naturally, the first 
question which arises in the mind of 
the mill operator is wherein will the 
use of electricity as a motive power 
lead to lower cost of production and 
a higher grade output. The answer 
to this question is found in a number 
of primary effects resulting directly 

shafts of driven machines with an 
average loss of from 20 to 25 per 
cent., from three to five per cent, of 
this being in the wiring and the bal- 
ance in the generators and motors. 
Moreover, we reduce the loss to this 
low point regardless of the com- 
plexity of the system or the distance 
of transmission. Distance enters into 
consideration only in so far as it 
affects the size of the wire required 
to transmit a given quantity of power. 
We can make the losses in the wire as 



from the use of the motor drive, each 
one of which bears its part in secur- 
ing the final result above mentioned. 
These primary effects may be classi- 
fied as follows : 

( 1 ) Reduction of transmission 
losses; (2) greater reliability; (3) 
greater elasticity; (4) reduced cost 
of maintenance ; (5) steady speed 
throughout the mill; (6) ideal condi- 
tions for the paper machine; (7) sav- 
ing in space; (8) decreased chance of 
injury to employees; (9) greater 

Production of Transmission Losses. 
— In every mill with old-style equip- 
ment the quantity of belting, pulley* 
hangers and shafting is something 
appalling, and the losses of transmis- 
sion through this complex system are 
very large. Actual tests show that 
these losses, which are usually termed 
friction losses, vary from 40 to 60 
per cent, of the power delivered at the 
engine or water-wheel shaft ; 50 per 
cent, being a very fair average. This 
figure, of course, does not apply to 
the grinders or the beater shaft when 
driven direct, as is usually the case. 
The more complex the system of 
shafting and bolting becomes, and the 
greater the distance of transmission 
from the source of power, the more 
rapidly does the percentage of fric- 
tion loss increase. In comparison 
with this we are able, with the elec- 
tric system, to transmit power from 
the shaft of the prime mover to the 

low or as high as we please by in- 
creasing or decreasing, respectively, 
the size of the wire. In practice, five 
per cent, is considered a commercial 
allowance for installations of this 

It is evident from the above that 
in the power transmission alone a 
saving of from 15 to 40 per cent., de- 
pending on the conditions to be met 
with in each individual case, can be 
effected by the use of the electric 
System. A modern two-machirfe mill, 
with the. prevailing tendency toward 
large size machines, will require out- 
side of the grinders, from 1500 to 
2000 h.p. Assuming the minimum 
saving of 15 per cent, on this, we 
shall have a net gain of from 22^, to 
300 h.p.. a very considerable sum 
when reduced to the basis of dollars 
and cents. In a steam-driven mill 
this saving would be represented by 
the decreased consumption of fuel ; 
in a mill where water is the prevailing 
power the saving will appear in the 
additional volume of water available 
for grinding pulp, or for extension - : 
and in these clays pulp is a very valu- 
able asset, which can always lie mar- 
keted to mills that are unfortunate 
enough not to possess sufficient water- 
power to grind their own pulp. 

The equipment of the new mill of 
the Watab Pulp & Paper Co., near St. 
Cloud, Minn., affords a striking 
example of minimized transmission 
losses. A visitor to this mill cannot 

help being impressed by the entire 
absence of the usual shafting and 
belting. The ]x>wer-hou^<.\ as it were, 
i- brought directly to each machine 
through the medium of a small wire 
on the ceiling, which requires prac- 
tically no space and is in no one'-, 

Greater Reliability. — Where the 
operation of any particular machine 
or line shaft i> dependent upon that 
of a number of other shafts, with 
many belts intermediate between this 
machine or shaft and the prime 
mover, there i> always the liability of 
an entire shut-down, due to the fail- 
ure of one of the intermediaries. In 
a paper mill this possibility is in- 
creased by reason of the fact that 
many of the belts are large and very 
heavy, causing a great strain t<> be 
put upon the shaft, hangers and pil- 
low-blocks and resulting in throwing 
the shafting out of alignment and 
heating the journal boxes, to say 
nothing of breaking the belts them- 
selves. It should not be forgotten, 
either, that a shaft out of alignment 
rapidly increases the friction losses. 

With the electric system there are 
no intermediaries. Each motor, with 
the machine or shaft which it drive-. 
is really a separate unit. and. as the 
failure of the wiring is a most un- 
usual occurrence, is dependent on 
nothing but the operation of the 
prime mover itself. In regard to the 
reliability of electric generators and 
motors, the best evidence of the high 
esteem in which they are universally 
held is found in the great and con- 


stantly increasing number of factories 
in all lines of industry which are 
operating under the electric system. 
The motor troubles, of which we 
occasionally hear, are in nearly every 
case directly traceable to poor engi- 
neering 1 in laying out the installation 
or to negligence on the part of those 
employed to look after the electrical 

Greater Elasticity. — The term elas- 
ticity, as applied to an electric system 
of power transmission, i> intended to 



February, 1908 

convey the idea of easy expansion and 
application to all of the many power 
problems encountered in the modern 
mill or factory. If the mill operator 
wishes to install new apparatus he 
does not have to stop to figure out 
how he is going to get power to the 

it is better to direct couple to the 
water-wheel shafts (although there 
are some enthusiasts who favor dri- 
ving these also by direct-connected 
motors). The equipment of the 
Watab Mill, previously referred to, is 
carried out along the lines recom- 



new machines — he knows that all he 
has to do is to install a motor and run 
a wire to it. It does not matter 
whether the new apparatus is up- 
stairs, downstairs, around the corner. 
01 across the street from that already 
installed ; it may be half a mile or 
more away, and still it is simply a 
case of a motor and some wire. Many 
times an entire new power plant is 
saved by this great elasticity. 

Another phase of this same elastic 
system lies in our being able to dis- 
tinguish between the different types 
of power service required throughout 
the mill and to treat each individual 
case on its merits and in the manner 
calculated to secure the highest effi- 
ciency of operation. We have con- 
stant speed motors for constant speed 
work, variable speed motors for vari- 
able speed work, elevator motors to 
drive the elevators and crane mot' >rs 
to operate the cranes. In the old 
system it was always a case of belt- 
ing: but with the electric system we 
can either direct connect, gear or belt 
the motor to the driven machine, as 
may seem best In general, for paper 
mills it is better to direct couple 
wherever this can be done without too 
great an investment in the motor by 
reason of the necessity for very slow 
speeds. The scheme usually followed, 
and which we recommend, is to 
direct connect the centrifugal pumps 
jordans, barkers, log- saw and vari- 
able speed shaft of the paper ma- 
chines : gear all duplex or triplex 
pumps, rotaries. calenders and wet 
machines : gear or belt the conveyors, 
rag dusters and cutters ; and belt the 
beaters, elevators and all group 
drives. The grinders, in our opinion, 

mended above, and the results have 
been most gratifying. 

Reduced Cost of Maintenance. — 
Every one who lias had experience 
with a complicated system of belting 
and shafting well knows what it 
means to keep it in running order. 

upon them by very heavy or very 
tight belts: shaft- thrown out of 
alignment by these same strains — in 
short, something continually happen- 
ing to claim the repair man's time and 
try the patience of the operator as 
well as draw on his bank account. 

O »ntrast the above with the situa- 
tion in an electrically equipped mill 
where shafting is practically elimi- 
nated. The chief engineer and his 
assistant can easily take care of the 
electrical apparatus, for all they have 
to do is to see that the bearings of 
the motors and generators are prop- 
erly lubricated, and, in case the equip- 
ment is of the direct-current type, to 
keep the commutators clean and the 
brushes properly adjusted. With al- 
ternating-current motors, which have 
no commutators, there is absolutely 
nothing to require attention but the 
bearings. Burn-outs in the windings 
of the motors are practically unheard 
of. but even if they should occur, the 
motor can be repaired in less time and 
with less expense than is required to 
restretch a broken belt or replace a 
melted-out journal. The services of 
at least one man and possibly two or 
three are saved outright and there is 
practically no expense for supplies, 
as none are required except for exten- 
sion- after the installation is com- 


In the average paper mill much of the 
time of the mill-wright and his assist- 
ants is required for this work. A 
broken or stretched belt, a short belt 
that refuses to pull right until it is so 
tight that the journals heat, belts that 
run off because they have to round a 
corner, and belts that wear because 
they are twisted : journals that heat 
because of the great lateral strain put 

Steady Speed Throughout the Mill. 
- — One of the many drawbacks to 
the belt and shaft scheme of trans- 
mission lies in the constantly chang- 
ing speed throughout the system. 
This change is not due so much to 
variation in the speed of the prime 
mover as it is to the effect of starting 
and stopping the various shafts or 
machines comprising the system. A 

February, 1908 



heavy load, thrown on or off in this 
way, reacts upon the apparatus near- 
est to it on the system, instead of 
upon the prime mover, as it should. 
With an electric installation, however, 
the effect of starting or stopping- any 
motor is carried back to the switch- 
board, which is connected directly to 
the prime mover, and, as the prime 
mover is always provided with a gov- 
erning device to control its speed, the 
other machines on the system are not 
interfered with. There are many ma- 
chines where a continued variation in 
speed becomes a serious matter, and 
at all such times such changes impair 
the running efficiency of the mill and 
tend to lower the quantity and quality 
of the output. 

[deal Conditions for the Paper Ma- 
chine. — The paper machine itself is 
the very heart of the mill, and the 
successful operation of the auxiliary 
apparatus is of no avail if there is 
trouble at this point. Nothing con- 
tributes so much to the ideal condi- 
tions for operation as a steady speed, 
which can at the same time be readily 
varied, and uniform temperature in 
the steam which is used for drying. 
Both of these requisites are possessed 
by the electric system. For reasons 
explained in the preceding section, 
we derive from the prime mover an 
almost absolutely constant speed. 
This speed, through the agency of the 
variable speed motor which has 
proven so satisfactory in service, can 
be varied at the will of the operator 
over a very wide range, and at the 

high as three or even four to one. 


the use of a controlling device and 
multiple voltage (although this slight- 
lv decreases the efficiency of opera- 
tion), that ratio of speed change may 
be still further increased to six, eight 
or even 10 to one. One of the most 

the motor is direct coupled. The 
speed can be varied over the entire 
range without a shutdown, which 
means that the machine tender can 
change from one grade of paper to 
another, no matter how great a dif- 
ference, without stopping the ma- 



chine. The difficulties of maintaining 
a constant speed, when the machine is 
driven from an engine operated un- 
der a varying steam pressure and 
varying load, are too well known to 
require more than mention. The 
clumsy and inefficient mechanical 
speed changers, step pulleys and 
cones are always a source of trouble 
and. compared with electric trans- 
mission are the makeshift of a pa>t 


The single objection raised by some 
manufacturers to the electric system 
is with regard to the economy of the 
drive on the paper machine itself. 
They contend that as long as it is 
necessary to use steam tor drying it 
would be better to drive the machines 
engines and use their exhaust 




same time the motor maintains prac- 
tically a constant efficiency. Variable 
speed motors on paper machines, in 
actual operation to-day. have a ratio 
of speed change of two to one under 
these constant conditions of efficiency, 
and it is entirely practicable to get as 



important features of the speed 
change lies in the fact that the suc- 
cessive increments of the change are 
very small ; each step on the con- 
troller means a change of only 1.5 
to two revolutions on the motor, or 
on the variable speed shaft to which 

rather than drive by motors and dry 
by live steam. In a consideration of 
this point we should divide the mills 
into two classes: First, those whose 
initial power is water, and second, 
those whose initial power is steam. 
We are certain that even the most ar- 
dent advocates of the use of steam 
engines and exhaust steam will ad- 
mit the superiority of electric drive 
in the case of mills of the first clas-^ 



February, 1908 

where ample water power is available 
for all purposes, for there can be no 
questions that the generation of elec- 
tric power from water is cheaper and 
more efficient than the generation of 
steam power from coal. The only 
argument, therefore, is with respect 
to mills of the second class, which 
must produce all or a part of their 
power from steam. Many such mills, 
partially steam driven, could be 
placed in the water-power list by the 
adoption of the electric system, the 
power thus saved being equivalent to 
that produced by steam. Admitting, 
for the sake of argument, that in mills 
of the second class it is more econom- 
ical to use engines on the machines — 
which contention is open to dispute — 

by water power, no engines whatever 
will be required ; if by steam, the en- 
gines will be used solely to drive the 
electric generators in the power- 
house. In either case the drying will 
be done by live steam taken directly 
from the boilers, and brought down 
ti ) the proper pressure through a re- 
ducing valve. By this method we are 
assured of a constant pressure, which 
means a uniform temperature in the 
dryers of the machine. There is, of 
course, no reason, in a steam-driven 
mill, why the exhaust from the en- 
gines driving the generators cannot 
be used for this purpose if desired, 
hut if the drying is done with exhaust 
steam, it is almost out of the question 
to maintain a uniform temperature, 



the many advantages of the electric 
drive, as described herein, are more 
than efficient to outweigh the possible 
saving in fuel. Moreover, we see no 
reason why the exhaust from the 
main engines driving the electric gen- 
erators could not be used for drying 
if desired ; and if this is done the 
question of economy is wiped out en- 
tirely. Another contention which has 
been advanced by paper manufac- 
turers who are using the electric 
drives on their machines and live 
steam for driving, is that the use of 
live steam permits of maintaining a 
constant temperature on the dryers, 
resulting in a uniform sheet of paper 
of high finish, a condition which can- 
not be as easily secured by exhaust 
steam on account of the varying pres- 
sure in the line. If the mill be driven 

as the exhaust frequently drops to so 
low a temperature that it is practically 
water and of little value for drying. 
Admitting, for the sake of argument, 
that it is cheaper to dry with exhaust 
than with live steam — which fact is 
open to dispute — the advantages of 
the live steam outweigh any possible 
saving in the use of the exhaust. 

Seizing in Space. — Doing away 
with so much shafting, with its ac- 
companying arrav of pulleys, belts, 
hangers and pillow-blocks, naturally 
relieves, to a great extent, the con- 
gestion which exists in the average 
mill. As the motors are mostly direct 
geared or coupled, very little space is 
required by them, thus permitting the 
placing of the various machines close 
together, and at the same time afford- 
ing ready access to all. As the re- 

sult, a large floor space is available 
for the storage of stock, and for stor- 
age in such a manner as to afford 
ready access, not only to the stock 
itself, but also to adjacent machinery. 
No greater contrast between electric- 
ally-driven and shaft-driven mills 
can be found than is evidenced by 
photographs of their respective base- 
ments. It is no uncommon sight, in 
mills of the latter class, to encounter 
^uch a mass of waste, stock and ma- 
chinery that it is almost impossible to 
get at either. 

Decreased Chance of Injury to Em- 
ployees. — The protection of the lives 
and limbs of his employees is, natu- 
rally, one of the first thoughts of a 
manufacturer, and yet hardly a day 
passes but that we read of some op- 
erative getting caught by a belt, pul- 
ley or shaft and either losing his life 
or being seriously injured. Such ac- 
cidents it is almost impossible to pre- 
vent where a system of shafting is in 
use — the two evils are co-existent. 
There is also a sequel to accidents of 
this character in the shape of damage 
stiits which, under the employers' lia- 
bility laws existing to-day, frequently 
results in large losses to the owners 
1 'f plants : in a recent case of this kind 
the plaintiff — a mill hand — received a 
verdict of $15,000. which was con- 
firmed by the Supreme Court. 

Greater Cleanliness. — In a paper 
mill cleanliness is of paramount im- 
portance, and nothing contributes 
more to a high-grade product than 
clean stock to a clean mill, particu- 
larly a clean machine room. Belts, 
pulleys and shafting are potent con- 
veyors and disseminators of dust, 
dirt, grease and filth of all kinds, and 
the only safeguard against these 
evils is to do away with pulleys and 
shafting, which can be accomplished 
only by substituting electric drive. 

In the foregoing we have endeav- 
ored to present, as briefly as possible, 
some of the more important argu- 
ments in favor of the electric system 
of power transmission for paper mills. 
The true significance of the state- 
ments can only be appreciated by 
those who have adopted the system, 
and of this number we feel safe in 
saying that there is not one who will 
not substantiate what has been said. 
One of the most encouraging features 
of the situation, and one which au- 
gurs well for the future expansion of 
electrical equipment in paper mills. 
lies in the fact that many operators 
who originally were skeptical or 
openlv opposed it are to-day loudest 
in sounding its praises. 

As an illustration of what can be 
accomplished by the adoption of the 
system advocated in the preceding 
paragraphs, a brief description of the 
new mill of the Watab Pulp & Paper 

February, 1908 



Co. may be interesting'. This mill is 
located on the Mississippi River, at 
Sartoll, Minn., a few miles above St. 
Cloud, and is one of the most modern 
and best equipped plants of its kind 
on the American continent. 

The mill buildings are all of rein- 
forced concrete and consist of a ma- 
chine room and basement 72. by 230 
ft., a beater room and basement 40 
by 122 ft., wood room and basement 
36 by 70 ft., boiler house 58 by 72 ft., 
pulp mill 32 by 152 ft., pump room 
20 by 70 ft., and power-house 32 by 
J2. ft. The boiler room contains two 
Sterling water tube boilers, which 


furnish steam for heating the build- 
ings and for drying the paper. All 
the power used in the mill, with the 
exception of that for the nine grind- 
ers, which are direct coupled to water- 
wheels, is generated in the power- 
house, which is located near the cen- 
ter of the river immediately adjoin- 
ing the grinding room. A dam 600 
ft. long was thrown straight across 
the river and affords a head of ap- 
proximately 17 ft.; the power-house 
and grinding room form a part of the 
dam, and the remainder consists of 
gates and spillway of the ordinary 
crib construction — rock filled. 

The generating equipment consists 
of three 600-kw., three-phase, 60- 
cycle, 480-volt, alternating-current, 
revolving field generators, each direct 
connected to a sextuplex water-wheel 
operating at a speed of 200 rev. per 
min. The exciting current for these 
generators is supplied from a 75-kw., 
1 20- volt, direct-current generator, 
coupled to a single water-wheel, or 
from a motor-driven exciter of the 
same capacity. Two motor-generator 
sets, each consisting of a 250-kw., 
240-volt, direct-current generator 
coupled to a 480-volt synchronous 
motor supply current to the variable 
speed motors in the machine room. 
The alternating system was adopted 
as being best suited to withstand the 
rigorous requirements of paper mill 
work and least liable to give trouble 
with motors when located in moist 
places, so common in every mill. The 
current is distributed through a 13 

panel blue Vermont marble switch- 
board, arranged with wattmeters and 
ammeters so that an accurate log may 
be kept of the power required by each 
separate department. Thirteen feeder 
circuits carry current to 51 constant 
speed alternating-current induction 
motors and to eight variable speed 
direct-current motors. 

The machine room contains two 
154-in. Beloit machines. The vari- 
able speed shaft of each machine is 
driven by a 250-h.p., variable speed, 
direct-current motor coupled directly 
in the center of the shaft. This mo- 
tor operates at a speed of from 270 
to 540 rev. per min., corresponding 
to a paper speed of from 300 to 600 
ft. per min. The entire variation is 
accomplished by means of resistance 
in the shunt field of the motor. The 
field rheostat controlling the speed of 
the motor has 138 steps, which means 
a change of approximately two feet 
per minute in paper speed for each 
step, which is surely a much finer 
variation than can be obtained 
through the medium of any mechan- 
ical speed changing device; it must 
be borne in mind, too, that the chang- 
ing of speed is accomplished without 
any loss of time whatever, as the con- 
troller is located on the machine room 
floor — about midway of the drying 
rolls— and the movement of the con- 

The only other direct-current motors 
in the mills are two of five-horse pow- 
er capacity with a speed range of two 
to one, which drive special cutters lo- 
cated in the machine room basement. 

The group at the wet end of each 
machine, in addition to the shake, 
consists of a save-all driven by five- 
horse power, back-geared induction 
motor, suction pump driven by 30- 
h.p., geared induction motor, fan 
pump with 40-h.p., direct-coupled mo- 
tor, machine pump with 7^2-h.p. 
geared motor and machine chest agi- 
tator and flat screens driven by 30- 
h.p. belted motor. To this group be- 
long 1 also the broke beater driven by 
40-h.p. belted motor and the broke 
beater pump driven by 40-h.p., direct- 
coupled motor. 

The beater room contains six beat- 
ers and two Majestic Jordan engines. 
The beaters are driven in pairs, one 
right and one left, by three 100-h.p. 
motors located in the basement. These 
motors have their shafts extended 
both ways with pulleys on either end ; 
outboard bearings are also provided 
to insure rigidity and freedom from 
belt slippage. Belts connect the mo- 
tors directly with the beater pulleys. 
The drive for the Jordans is one of 
the most noteworthy in the mill. Each 
Jordan is directly coupled to a 150- 
h.p. motor operating at 312 rev. per 




Horse Power 


Machine Driven 



















Log Haul 

Conveyor and Splitter 

Prepared Wood Conveyor 

Wood Room > 

Log Saw and Live Rolls 


Shaving Fan 

Coal Conveyor 

Pulp Pumps 

Decker Pumps 

Pump Room 

Pressure Pump on Grinders 

Back-off Pressure Pump on Grind' 

Water Pumps 

Sliver Screen 

Wet Machines 

°° m " } 

Flat Screens and Deckers 

Machine Shop 

Power-House > 


Lathes, Planers, etc. 

Dir. Current Generators 
Dir. Current Excitor 

Tordan Engines 


Beaters . 


Stock Pumps 


Stock Chest Agitators 

Dir. Cur. Var. Speed . 
« « 

Var. speed Shaft of Machine 



Suction Pumps 

Machine Room , 


Fan Pump 
Broke Beaters 


Broke Beater Pumps 


Machine Pumps 


Agitator and Screen 



Finishing Room 

Dir. Cur. Var. Speed . 


troller handle is but the work of a 
second for the machine tender. The 
re winder at the head of each machine 
is also driven by a variable speed of 
30-h.p., direct-current motor having a 
speed range of two to one. Two 
7^2 -h. p. motors of the same type, 
with 25 per cent, speed variation, op- 
erate the shakes of the two machines. 

min. The base of the Jordan is ex- 
tended to receive the motor, which is 
arranged to slide on rollers as the 
plug is moved in or out by means of 
the regulation hand wheel. In the 
beater room basement are located the 
stock pumps for supplying the Jor- 
dans, each driven by a 7 T j-h.p. geared 
motor and also the Jordan chest agi- 



February, 1908 

tators driven by two 15-h.p. belted 
motors. The controllers for starting 
and stopping- all motors governing 
the supply of pulp to the beaters and 
Jordans are located on the beater 
room floor, so as to facilitate rapid 
handling of stock. 

In the pump room are located the 
pumps controlling the water supply 
and those for handling the ground 
wood pulp. The entire water of the 
mill, outside of the boiler feed, is ob- 
tained through three six-inch centrifu- 
gal pumps, each direct connected to 
a 40-h.p. motor. An underwriter's 
pump is used for fire purposes only. 
Two centrifugal pumps for pumping 
pulp from the grinders to the flat 
screens in wet machine rooms are lo- 
cated in a pit beneath the pump room. 
Each of these is direct connected to a 
50-h.p. motor. In this room are also 
the centrifugal Decker chest pump 
direct coupled to a 40-h.p. motor, the 
triplex pressure pump for grinders 
driven by a 40-h.p. geared motor, the 
triplex back-off pressure pump driven 
by a 73>2-h.p. geared motor and the 
silver screen driven by a five-horse 
power back-geared motor. 

The wet machine room contains 
five wet machines, two deckers and 

12 flat screens. The wet machines 
are each driven by a 10-h.p. back- 
geared motor. The flat screens, 
which are located immediately back 
of the wet machines, are driven 
through friction clutches connected to 
a line shaft by means of a miter gear, 
this line shaft being direct coupled to 
a 100-h.p. motor. The deckers are 
driven by belt from the wet machine 
shaft. A machine shop adjacent to 
the wet machine room is driven by a 
10-h.p. belted motor. 

The wood room equipment consists 
of six barkers and a 60-in. log saw 
with the necessary conveyor. The 
logs used for making pulp are taken 
from the river to the saw by a log 
haul which is driven by a five-horse 
power hack-geared motor. The saw 
and live rolls are direct coupled to a 
30-h.p. motor, the coupling being of 
sufficient weight to act as a fly-wheel, 
and thus reduce the actual power re- 
quired for sawing. The sawed wood 
i- carried to the barkers by a conveyor 
driven by a 10-h.p. back-geared 1110- 
tor; tin- same motor also drives a 
splitter. The barkers are each driven 
by a 7 r j-h.p. direct-coupled motor, 
the iron base of barker being cast 
with a pedestal to receive the motor. 

The prepared wood is carried to the 

grinding room by a conveyor driven by 
am >ther 10-h.p. back-geared motor. A 
shaving fan driven by a 40-h.p. belted 
motor blows the chips from the bark- 
ers directly on to the boiler grates 
What additional coal fuel is required 
is carried in hoppers above the boilers. 
A 10-h.p. back-geared motor drives a 
bucket elevator which fills the hop- 
pers directly from the cars in which 
the coal is shipped. 

The Watab mill has now been in 
continuous operation for nearly one 
year. It began making paper a few- 
hours after current was first turned 
on. and has not been shut down since 
due to failure of equipment — a most 
unusual performance for a mill of this 
size. The complete electrical instal- 
lation was furnished by the Allis-Chal- 
mers Company, and the design and 
specifications were prepared by engi- 
neers of the Allis-Chalmers Company, 
acting in conjunction with the mill 
superintendent and Mr. Samuel 
Holmes, paper mill expert and engi- 
neer. The officers of this mill are 
justly proud of their property and 
are ready and willing at all times to 
answer inquiries from interested par- 
ties pertaining to the mill and its 

The Westing'House Electrically- 
Heated Sad-iron 

In form the Westinghouse Electric 
iron differs only from an ordinary 
iron in its more symmetrical and at- 
tractive appearance, and in being pro- 
vided with terminals and a flexible 
cord through which the current is 
transmitted. In operation it elevates 
ironing from hot drudgery to a com- 
fortable and pleasant task. Its heat- 
ing mechanism, which is entirely con- 
cealed, keeps the iron at the proper 
temperature at a minimum consump- 
tion of current. It consists of a flat 
insulated resistance strip clamped by 
hydraulic pressure between two flat 
iron plates, forming a solid heating- 
element of high thermal conductivity, 
and having a large heat storage ca- 

The design of the element is such 
that the heat is evenly distributed 
over the entire bottom of the iron, 
the edges and point being practically 
at the same temperature as the middle 
of the bottom. A non-conducting ele- 
ment is used between the top of the 
iron and the heating unit. This con- 
struction results in the top of the iron 
being cooler than its face. The heat- 
ing- unit is hermetically sealed in its 
insulation and cannot, therefore, de- 
teriorate any faster than the iron, as 
it is not subject to the oxidizing ef- 
fects caused by contact with air. 

The iron is extremely • simple and 

there is absolutely nothing to get out 
of order or to require renewal. Irons 
which have been in use continually 
for the past six years, without cost- 
ing a cent for repairs, are in as good 
condition to-day as ever. In fact, the 
electrical construction is so simple and 
durable that there is no reason why it 
should not last as long as the iron 

The terminals are protected by a 
solid metal guard, and the cord lead- 
ing from them is securely anchored 
by a clamp on the handle of the iron. 
A wire spring which surrounds the 
cord at this point prevents any sharp 
bends. A separable plug is provided 
by which the iron can be connected to 
any convenient lamp socket without 
twisting. A small spring attached to 
the cord takes up all slack and so 
keeps it from dragging on the clothes 
or getting in the way of the operator. 

The polishing surface of the iron 
is of highly polished cast iron, found 
by experience to be the most satis- 
factory surface for the purpose. The 
upper portions of the iron have a bur- 
nished nickel-plate surface. The han- 
dle is of ebonized wood, heat-proof 
and unbreakable. A heat-proof stand, 
upon which the iron should be set 
when not in use, is provided with 
every iron. 

The irons arc made to suit all com- 
mercial lighting circuits, and can be 
used on either alternating or direct 

current with equally satisfactory re- 

The Electric Drive in a Hardware 

The popularity of the electric motor 
drive is due in no small part to the 
general satisfaction that has attended 
the use of squirrel cage induction 
motors. Simple and rugged in con- 
struction, reliable and efficient in 
operation, they are remarkably free 
from the operating difficulties that at- 
tend the use of the more sensitive 
commutator motor. Manufacturers, 
realizing these as well as the further 
advantages pertaining to the use of 
the electric motor drive as a whole, 
now specify electric motors for power 
purposes when contemplating the 
erection of a new plant or the enlarge- 
ment of the old. 

A notable example of the applica- 
tion of the induction motor drive to 
the solving of power problems in a 
modern manufacturing establishment 
is found in the factory of the O. M. 
Edwards Company, of Syracuse. X. Y. 
This company is an extensive manu- 
facturer of hardware specialties used 
in both the steam and electric rail- 
road trade, such as window and ex- 
tension platform trap-door fixtures. 
steel window sash, molding, etc. For- 
merly, the company purchased power 
from the Syracuse Lighting Co.. us- 
ing 500-volt direct-current motors for 

{.Continued on page 10 of ad. section). 


Volume XXXIX. Number 3. 
$1.00 a year; 15 cents a copy 

New York, March, 1 908 

The Electrical Age Co. 
New York and London 

Published monthly by The E'ec'.rical Age Co., 45 E. 42d Street, New York. 

J. H. SMITH. Pres. C. A. HOPE. Sec. andTreas. 


Telephone No. 64'J8 38th. Private branch exchange connecting all departments. 
Cable Address — ~Revolvable, New York. 


United States and Mexico, $1.00; 

Canada, $1.50. To Other Cointries, $2.50 


Insertion of new advertisements or changes of copy cannot be guaranteed for the following 
issue if received later than the 15th of each month. 


Business and Editorial Departments, 33 Bedford Street 

Strand, W. C. 

London Telephone, Post Office, 3807 Central. Private branch exchange. London Cable 

Address Revolvable, London. 

All remittances to London Office should be crossed London and County Banking Company. 


Subscriptions may be sent to the following branch offices or agencies, where they will re- 
ceive the same careful attention as at the home office in New York : 

124, Queen St., Melbourne 33, Loop St., Cape Town 

359, George St., Sydney Unter den Linden, 5, Berlin 

61, Main St., Yokohama Nevsky Prospect. St. Petersburg 

23, Esplanada Rd., Bombay 31. bis rue de Fauburg Monlmartre, Paris 

General Agents for United States and Canada : The American News Company 

Illuminating Engineering 

"Illuminating engineering is not a 
matter of light distribution — it is a 
matter of suitable lighting, and the con- 
ditions determining what is suitable 
are just as different as any two de- 
signs that are different," was the state- 
ment of Bassett Jones, Jr., before the 
Illuminating Engineers' Society on 
December 12, 1907. 

We are inclined to believe that Bas- 
sett Jones, Jr., is right. The matter 
is very plain. Architects have always 
fixed the lighting of the edifices they 
design and always will. Lighting is 
only a detail, but it is a very im- 
portant one for all that. The trouble 
about it arises out of the confusion 
of new illuminants. 

We are quite prepared to admit that 
it would absorb the entire energies 
of an individual mind to post itself on 
the changeful types and the evolution 
of particular forms of electrical light- 
ing. Assuredly, there is room at the 
present for expert knowledge in the 
matter, but is it likely to continue 
longer than the settlement of manu- 
facturer's claims for illuminants? 

Owing to the rapid spread of elec- 
tricity as an illuminant, and the de- 
sire of large commercial interests to 
further that object, there has been cre- 
ated almost by a fiat a branch of en- 
gineering known as illuminating en- 
gineering. We are of the opinion that 
it will not long maintain a separate 
existence, and that its name will not 
endure as a distinct craft. While un- 
doubtedly the introduction of a vari- 
ety of electrical illuminants with vary- 
ing degrees of merit and varying 
limitations, has caused much concern 
to architects, and unquestionably cre- 
ated a demand for a man who can 
definitely regulate the light distribu- 
tion of units, there is no warrant to 
believe that the 'so-called profession 
of illuminating engineering will have 

a long tenure or prove attractive as 
a life work. 

The architects themselves are begin- 
ning to understand. They are now 
training men to study the technical 
limits of the lights which they use. 
In a few months an ordinarily trained 
mind can familiarize itself with the 
laws governing the distribution of 
light, and the facts about the quality 
of commercial illuminants. 

When it is clearly understood that 
knowledge of this sort is a desirable 
part of the equipment of an archi- 
tect, it is only a question of time until 
schools of architecture get definite 
courses on the subject and text-books 
contain adequate instruction on the 
use of illuminants. While we realize 
that prophecy is a dangerous art, we 
nevertheless predict that after a few 
years the matter of illumination will 
not be a problem for electrical experts 
at all, but a matter for the regulation 
of architects. 

By years of training the architect 
becomes intuitively familiar with the 
use of colors in decoration and with 
the esthetic distribution of lighting 
fixtures. It is impossible that an en- 
gineer will ever get this intuitive 
knowledge, and so it is that urging 
engineers to "study" this sort of thing 
is like carrying coals to Newcastle. 

New York City probably contains 
more engineering attempts at "placing 
light" than any other spot in the coun- 
try ; many of them are creditable, some 
are distinctively wretched. No more 
miserable work can be cited than the 
illumination of the United Engineer- 
ing Society's Building. In its library on 
the thirteenth floor one has a feeling 
that he is groping in the dark among 
books and stacks, and outlines of 
tantalizingly shadowy form, and he 
emerges into the corridor straining his 
eyes to see. It is a welcome relief to 
leave it behind. 

Non-Synchronous Generators 

The paper on non-synchronous gen- 
erators by Mr. W. L. Waters, read be- 
fore the American Institute of Elec- 
trical Engineers, at the February 
meeting, is very suggestive and should 
be carefully considered by all central 
station men. Mr. Waters proposes the 
use of the well-known polyphase in- 
duction motor as a generator. V\ 'hen 
such a motor is connected to aline on 
which electromotive force is already 
maintained and an engine drives its 
rotating part, it will revolve slightly 
faster than synchronism and the elec- 
tromotive force generated by the ro- 
tation will be somewhat greater than 
that of the line, and current will be 
supplied by the motor. 

This type of generator has sev- 
eral characteristics which are radically 
different from the usual or syn- 
chronous type. Its salient character- 
istics are as follows : 

It does not have to be synchronous 
at all, but may be thrown on the line 
at any time, provided its speed is ap- 
proximately right. This fact results 
from the machine having no field of 
its own. It takes its magetizing cur- 
rent from the line just as when used 
as a motor. For this reason there 
must be voltage on the line before the 
motor, or rather the generator, can be 
thrown on, or it will generate no cur- 
rent, since it will then have no mag- 
netism in the iron. 

When a line is short-circuited it will 
not add current to the short-circuit 
current, since the killing of the line 
voltage kills the magnetism of the 
generator. It will not, therefore, add 
to the short-circuit current, which is 
so destructive in many cases — this is 
rather an important consideration. 

The magnetizing current which is 
taken from the line by the non-syn- 
chronous generator is, of course, a 
lagging current. Furthermore, the 




March, 1908 

nature of this generator is such that 
it can supply current only in phase 
with the line voltage. 

That the non-synchronous genera- 
tor cannot supply lagging current is 
most important, and is one of the 
characteristics of this particular type. 
When any lagging load is to be car- 
ried on a system in which this sort of 
generator is used, all of the lagging 
component must be carried by the 
supply that maintains the line voltage 
initially for magnetizing the non-syn- 
chronous generator. It is the great- 
est disadvantage of the new type of 
generator and is serious, for genera- 
tors of the usual synchronous type 
must then operate under unfavorable 
conditions at a relatively low power- 

The new type of generator is espe- 
cially useful with steam turbines, for 
the short-circuited secondary may be 
made the rotating part, and any con- 
venient mechanical construction used 
for withstanding the centrifugal 
strains, since no insulation need be 
here maintained. Collector rings are 
also avoided, which is of some impor- 
tance in turbo-generators. 

Mr. Waters states that these ma- 
chines when used as turbo-genera- 
tors will often have a higher efficiency 
than the old type, and also a very 
high power-factor. Though he makes 
no actual statement as to cost, he 
leaves the impression that they will 
not be at a disadvantage, at least in 
many places. 

In an innovation as radical as this 
it is impossible to assign the actual 
field of usefulness in advance of trial 
and experience and the detailed lay- 
ing out of particular plants, but on the 
other hand the advantage of having a 
generator that does not have to be 
synchronized, which will not give 
enormous currents on short circuit 
and which is so mechanically satis- 
factory for turbine work should cause 
engineers to give it very careful 

Mr. Waters suggests a number of 
uses for the machine and outlines a 
number of ways in which its great 
weakness, the need of an additional 
generator to excite and to carry the 
lagging load may be overcome. His 
argument is not in all cases convinc- 
ing, especially where he advocates ob- 
taining of the exciting current from 
the substation rotary converters. 

Large central stations already op- 
erating and needing to add a large 
amount of additional power can well 
afford to consider whether a non-syn- 
chronous turbo-generator unit would 
not be the most satisfactory type. Un- 
der these conditions the lagging cur- 
rent could readily be carried by the 
old apparatus and a large block of 

power added in the non-synchronous 
unit when desired. The great simplic- 
ity of this type, the elimination of syn- 
chronizing and the great overload ca- 
pacity are important considerations. 

It is worth noting that the new type 
of generator has a very strong ten- 
dency to smooth out all harmonics 
from the line wave, which is a ma- 
terial advantage, though one easily 

A rather ingenious use of the new 
machine is cited by Mr. Waters in a 
plant delivering heavy low voltage di- 
rect current at some distance from the 
source of power. A non-synchronous 
generator drives a rotary converter at 
the end of the line, and ah of the ex- 
citing current is supplied by the rotary 
converter, which must be first started. 
Xo synchronizing is required. This 
plant has operated very satisfactorily. 

Standard HandbooK for Electrical 


Another handbook has come, will 
live its short life and will then be 
put on the shelf where the out-of- 
date books are kept, and all this 
will happen before its fine red 
cover shows signs of age. Mr. Otis 
Allen Kenyon deserves much credit 
for the painstaking way in which he 
has collected an immense amount of 
data and presented it in a rational 
and consecutive order, but it is to be 
regretted that he did not use his am- 
ple knowledge of the English lan- 
guage to present it in a more con- 
densed form better available for ref- 

While the typographical work is 
excellent, except for its small size, the 
plates are anything but first class. 
The occasional use of heavy type to 
catch the eye i- excellent and is car- 
ried out consistently. 

There is one fault which has been 
handed down like an heirloom from 
handbook to handbook which this 
book suffers from more than any that 
has gone before. It tries to tell you 
all about machine design — and fails. 
Machine design is the work of experts 
employed by manufacturers ; they are 
supposed to know all that has been 
published on their subject and in ad- 
dition know many trade secrets and 
tricks which, as a rule, only leak out 
of the factories when they are ob- 
solescent. Then after going the 
rounds of the engineering societies 
and technical papers, they find their 
way into the handbook compilers' file 
and in due course of years appear in 
the handbook. Authors should learn 
from the old story of Atlanta not to 
lose the race by stopping to pick up 
the discarded golden apples. The 
first 65 pages of Section 7 on Elec- 
tric Generators could, with advantage, 
be reduced to five pages. The re- 

mainder of the section is compre- 
hensive and clear, but the material is 
spread out rather thin. Sections 6 
and 8 are also attempts to tell ma- 
chine designers how to design ma- 
chines (for who else could use the 
information?); but in the interest of 
our electrical industries we hope that 
the book is not needed for that pur- 

Section 11, on Transmission and 
Distribution, is one of the most im- 
portant by reason of the fact that 
most of the matter treated therein is 
not in the hands of experts continu- 
ously engaged on the work, and is 
therefore a section to be often re- 
ferred to by those who have grown 
rusty. It is unfortunate that this sec- 
tion is very badly arranged for refer- 
ence and contains serious errors. 

Referring to page 699, Xo. 213, 
there is a list of wire gauges, the 
"chief ones" only being mentioned. 
The B. & S. and B. W. G. are not 
among these ! 

At page 704, Xo. 237, the reader 
is recommended to cover a copper 
ground plate with coke. It is per- 
fectly well known that a copper plate 
under such conditions will last for a 
few months only on account of the 
action of the sulphur contained in the 

Page 706, No. 246 gives the cost 
of stringing railway feeders from 
$30 to $60 per mile. The upper limit 
is very low and should, in fact, be 
more than doubled. 

Referring to page 637, X T o. 26, the 
following statement occurs in a cal- 
culation : 

Impedance per wire = 

Line drop per wire _ 2500 

Current per wire 235 

where 235 is the current and 2500 is 
defined as half the difference between 
transmitted and received voltages. 
The impedance is most emphatically 
not the ratio of half the difference 
between transmitted and received 
voltages to the current. It is the ra- 
tio of the potential drop in the line to 
the current, quite a different matter if 
the line power-factor differs from the 
load power-factor. 

A technical error of this character 
carried through numerous examples 
covering several pages is apt to make 
one suspicious of the whole work, a 
feeling somewhat intensified when 
page 907 in the section on Electric 
Traction is reached, and the table en- 
titled "Labor" shows a total of $903.- 
75 at the foot of a column which adds 
up to $488-75- 

Although the section on Traction is 
carelessly and incoherently written, it 
contains a wealth of data for the pa- 
tient man who has time to look for it 
and separate it from the catalog stuff 
with which it is surrounded. 

March, 1908 



Telephony and telegraphy are not 
treated at great length, and one won- 
ders why they are treated at all, be- 
ing foreign to the rest of the book, 
which really deals with power engi- 

Section 19 is mostly taken from an- 
nual publications. 

One of the most extraordinary col- 
lections of errors is that to be found 
concentrated on page 895 in the table 
on Third Rail Data. 

While nobody thinks the B. R. T. 
Co. has their rail installed to within 
one-quarter inch, their official third 
rail gauge is 203/2 in. from center of 
third rail to center of track rail, which 
is equivalent to 21^4 m - from the 
gauge line of the track rail. Under 
the circumstances, the "Standard" 
may perhaps be forgiven for stating 
22 in. instead of 21^/4. More serious 
is the error in the third rail gauge of 
the Interborough, given as 22 in. in- 
stead of 26 in., as correctly given in 
the New York Electrical Handbook, 
p. 296, McGraw Pub. Co., and in 
Street Railway Journal, March 4, 
1905, p. 427. The height of third 
rail surface is given in the "Standard" 
as 4^2 in., while both of these refer- 
ences correctly agree that the proper 
value is four inches. The gauge for 
the Long Island R. R. is given as 
2j l / 2 in., and should be 2j in., as 
stated in the Street Railway Journal, 
p. 832, vol. xxvi No. 19. The gauge 
for the West Short R. R. is given as 
2834 in-, and should be 32 in., accord- 
ing to the Street Railway Journal, vol. 
xxix No. 23, p. 1004. Again the 
gauge of West Jersey & Seashore, 
which ought now to be called the 
Camden & Atlantic City Ry., is given 
as 27^2 in., while the authority of the 
Street Raik<'ay Journal, vol. xvii No. 
19, p. 938, is in favor of 26 in. The 
Philadelphia & Western is said to 
have its third rail face six inches 
above the top of track rail. The Street 
Railway Journal, vol. xxix No. 24, p. 
1055, shows that it is 3^5 in.. 

We have not had sufficient time to 
check all figures, but having found 
seven errors in checking about a 
dozen cases, it would seem the part 
of wisdom to use this table with cau- 

It is an excellent idea to present a 
composite clearance diagram, but it 
would, have, been better if the clear- 
ance diagrams of the New York Cen- 
tral and Pennsylvania Railroads had 
been consulted in its preparation. As 
it stands, . the height is about nine 
inches shy and the width about four 
inches shy. Mr. Armstrong's com- 
posite diagram will fit loosely into 
those of the New York Central or 
Pennsylvania Railroads. 

On page 189 is an alleged quotation 
from The Electrical Age, of April, 

1907, relative to rubber insulation. 
The minimum desirable insulation re- 
sistance is quoted as 177 by io 7 meg- 
ohms per cm. 3 Reference to the Age 
of that date shows the original figure 
to be 750 by io G megohms per inch 
cube. This reduced to cm. 3 is 190 by 
io 7 , not 177 by io 7 , so that here is 
plainly a culpable error of calculation. 

House Bill No. 10457 

The acquisition of national forests in 
the Southern Appalachian Mountains 
and in the White Mountains is pro- 
vided for in House Bill No. 10457, 
now before the House of Representa- 
tives for consideration. There is con- 
siderable opposition to the bill, rising 
chiefly from two sources : timber in- 
terests and the political constituents of 
the mountainous sections covered by 
the bill. The enforcement of forestry 
regulations by the Government would 
enhance the cost of timbering and re- 
quire the education of lumbermen ; this 
were difficult, hence opposition from 
this source. The mountainous poor 
have always been accustomed to use 
the forests as they pleased, and Gov- 
ernment supervision means a curtail- 
ment of what has been regarded as 
the natural liberty of this class of peo- 
ple. In many sections of the South 
the wdiites living along the fringes of 
timber land are a considerable factor; 
hence another source of opposition to 
this bill. Nevertheless, it remains 
plainly to be seen that steps must be 
taken to protect our forests from fur- 
ther destruction ; a comparative tim- 
ber famine is even now upon us. 

The interest of engineers and in- 
vested capital in water-power projects 
lies in the preservation of stream flow 
in the rivers, streams and creeks 
which are actually or potentially valu- 

The power of the soil to equalize 
the flow of water forming upon it is 
well known, and the protection of soil 
at the water sheds can be attained 
only by proper forest regulations. 
The soil of forest land is not usually 
deep, and is rarely over a few inches 
in depth on sloping land. 

The soil consists of two parts more 
or less intimately mixed, the soil 
proper, formed by decomposition by 
the elements of underlying rocks, and 
the humus, or decayed vegetable mat- 
ter formed by the imperfect oxidation 
of woody material. 

Soil is worn away or denuded from 
the rock by the continual washing of 
rain driven at usually high wind ve- 
locities ; the trees themselves break 
the cutting power of the rain and the 
moisture drops upon the soil, being 
absorbed by the spongy humus, from 
which it is gradually released. The 
presence of the trees serves another 

purpose, viz. : the prevention of the 
evaporation of the water contained in 
the humus and thus further tends to 
prolong the flow of water from the 
soil. We must emphatically dissent 
from the opinion of an esteemed con- 
temporary, which states that "even the 
greediest timber hogs could not cut 
"ii watersheds of most streams fa»t 
enough to permanently damage by the 
mere removal of trees." What the 
author of this statement really intends 
to say is that the proper removal of 
trees, even to the point of barrenne-^. 
would not materially damage the 
humus, and that with the preservation 
of the humus the growth of second 
timber would begin. 

The latter process is frequently re- 
tarded and even completely destroyed 
by the entire destruction of the hu- 
mus. Tangles of brush remaining 
after the trimming of trees quickly 
dry out and are often annoying in the 
clearing of the forests. Careless lum- 
bermen remove this cutting debris by 
fire, and as the burning continues for 
days and is frequently little watched, 
it as often spreads to virgin forest 
timber. As a result of the fire we 
have imperfect combustion at the base 
of the flame-swept area and the re- 
duction of vegetable matter to carbon, 
the process extending wholly through 
the humus of the soil and completely 
destroying its fertility. The process 
is very like that under which vegetable 
charcoal is manufactured, but it takes 
place on a larger scale. With the 
complete destruction of the soil, there 
is a checking, if not an actual cessa- 
tion, of the process of regrowth. 

Speaking of this condition in 1896, 
the late Professor N. S. Shaler. of 
Harvard University, the foremost ge- 
ologist of this country, said : "South of 
Pennsylvania there is, according to 
my reckoning based on observation in 
every State in that upland country, 
an aggregate area of not less than 
3000 square miles where the soil has 
been destroyed by the complete re- 
moval of the woods and the conse- 
quent passage of the earthy matter to 
the lowlands and to the sea. At the 
rate at which this process is now go- 
ing on, the loss in arable and forest- 
able land may fairly be reckoned at 
not less than 100 sq. miles per annum. 
In other words, we are each year los- 
ing to the uses of man. through un- 
necessary destruction, a productive 
capacity which may be estimated as 
sufficient to sustain a population of a 
thousand people. This rate has not 
only been kept up, it has been greatly 
accelerated. Faster than was consid- 
ered possible 1 1 years ago. these re- 
gions, through injudicious cutting, 
fires, clearing and general misappro- 
priation, are moving toward a forest- 
less, soilless condition." 

The Central Station Distributing System 

Pressure Regulation 

THE inherent properties of the in- 
candescent lamp which make it 
highly sensitive to variations in 
pressure necessitate refinements of 
regulation in electric lighting work 
which are not required in purely power 
or traction enterprises. As the excel- 
lence of a central-station's lighting 
service is determined very largely by 
the care and attention given to pres- 
sure regulation, much thought has 
been given to this subject by engineers 
from the earliest days of the industry. 

In general, the regulation of pres- 
sure is accomplished by variation of 
bus voltage and by controlling the 
pressure on individual feeders by 
means of feeder regulators. 

In direct-current networks uniform 
pressure is maintained on the mains by 
varying the bus pressure as the load 
changes, and by the fact that the regu- 
lation of pressure is, to some extent, 
automatic. When a heavy load is 
thrown on at any point in a network 
the pressure near that point is reduced, 
causing current to flow from all ad- 
jacent feeders toward the low point in 
proportion to the capacity of the mains 
in the vicinity of the load. The heavy 
load is thus carried in part by all the 


Commonwealth Edison Co., Chicago 

lower pressure at remoter points. In 
the earlier development of networks 
it was customary to insert resistances 
in the feeders to prevent the pressure 
from running too high and thus rob- 
bing the longer feeders of a part of 
their load. In some cases this is un- 
avoidable in modern practice, but the 
loss of energy in the feeder resistance 
is a considerable item, and the space 
required for their installation is 
usually prohibitive where feeder loads 
are heavy. 

It is, therefore, found desirable to 
provide two or three separate bus bars, 
and arrange the switchboard so that 
the shorter feeders can be carried on 
one bus, those of medium length on 

plicable to stations and substations 
having several units. 

The operation of several busses is 
necessary only during the hours of 
heavy load, since the difference be- 
tween loss of potential on the longer 
and shorter feeders is not so great dur- 
ing the hours of light load, and all 
feeders can be carried from one bus. 

It is also possible to prevent pres- 
sure from running too high during the 
light-load period by opening a part of 
of the feeders running to a given sec- 
* tion, thus increasing the drop to the 
ends of the remaining feeders. 

It is necessary in some cases with 
very long feeders to install a motor- 
driven generator called a "booster" in 
series with the feeder to hold its pres- 


7o F<se&e/- 

r~ / YYYYYYYY N ] A i 


FIG. I. 

o 1 

FIG. 2. 

feeders nearest the low point, and this 
tends to prevent the pressure from 
falling as low as it otherwise would. 
( )n the other hand, the adjacent feed- 
ers being given more load, the pressure 
at their ends is lowered and the sys- 
tem tends thus to automatically equal- 
ize the pressure throughout. 

The different lengths and conductiv- 
ities of the longer and shorter feed- 
ers tend to produce higher pressure 
mi (Re network near the station and 

6 4 

another bus and the longest on a third 
bus. Each bus is supplied by genera- 
tors, rotary converters, or batteries, 
which can be independently regulated 
and each zone may, therefore, be car- 
ried at a pressure suited to the drop 
on its feeders without the use of feed- 
er resistances. This arrangement, of 
course, requires a sufficient number of 
sources of supply of the proper ca- 
pacity to carry the loads on the sev- 
eral buses, and is, therefore, only ap- 

sure up. Such boosters are usually 
overcompounded to automatically 
maintain constant pressure at the feed- 
er end at all loads. Where storage 
batteries equipped with end-cell 
switches are available it is sometimes 
feasible to put such feeders on the bat- 
tery through a separate bus and thus 
avoid the use of a booster. The use 
of a booster is not to be recommended 
on long feeders until the cost of feeder 
copper required to produce equivalent 

March, 1908 



results considerably exceeds the cost of 
the booster equipment. 

It is usual in low-tension networks 
to run pressure wires from the prin- 
cipal feeder ends back to the station, 
where they are connected to a multiple 
point switch in such a way that a 
voltmeter may be connected to the 
pressure wires of any feeder, and the 
pressure at any point in the network 
may thus readily be known at any 

fig. 3. 

In operating the system a feeder 
which represents the average condi- 
tion in any zone is selected as a 
"standard feeder." The pressure wires 
of this feeder are run to a separate 
voltmeter, which is used for regulat- 
ing the bus which supplies the zone. 
The operator manipulates the field 
rheostats of the machines which are 
operating the bus, as may be necessary 
to hold the pressure as indicated by 
the voltmeter on the standard feeder 

A similar standard feeder is re- 
quired for each bus, and in large sys- 
tems a second standard is sometimes 
selected for use in case of emergency. 
In stations where a storage battery 
auxiliary is provided it is usual to ad- 
just the battery pressure to that of the 
bus and to connect them in parallel. 
This permits the battery to "float" on 
the bus and to automatically charge 
and discharge as the pressure rises 
above, or falls below, the normal. 
The effect of this is to steady the bus 
pressure greatly and to maintain it in 
case of interruption of the power sup- 
ply, provided the load does not ex- 
ceed the limit of battery capacity 
when the interruption occurs. 

In alternating-current systems the 
problem of pressure regulation is 
solved in quite a different way. In- 
dividual feeder regulators, which 
waste but little energy, permit the 
economical operation of all feeders on 
one bus, if it is desired. In low-ten- 
sion networks the conditions are very 

similar to those found in a direct-cur- 
rent network, but the problem is much 
more easily met because of the avail- 
ability of feeder regulators. These 
regulators are also found useful «in 
relieving overloaded feeders when the 
mains interconnecting them are of suf- 
ficient capacity to permit it. When a 
feeder becomes overloaded the regula- 
tors on adjacent feeders may be used 
to raise these feeder-end pressures, and 
this causes these feeders to take part 
of the overloaded feeder's load. 
Where the feeders are low tension, 
as well as the mains, and are installed 
underground, pressure wires may be 
embodied in the feeder cables, as is 
customary in direct-current distribu- 
tion, at a small expense. If the lines 
are overhead, or the construction is 
such that separately insulated pressure 
wires are required, it is usually less 
expensive to utilize line-drop com- 
pensators instead. 

In areas in which the load is so scat- 
tered that the distribution is effected 
chiefly by means of primary mains, it 
is usually found desirable not to inter- 
connect adjacent feeders. This re- 
quires that each feeder be inde- 
pendently regulated to deliver the 
proper pressure at its terminus and 
feeder regulators are, therefore, very 
essential to a system having a num- 
ber of feeders of different lengths and 
sizes. The design of an efficient and 
practical form of feeder regulator is 
fortunately not difficult, and there are 
two types in general use in America. 
Stillwell at an early date devised a 
transformer with a secondary winding 
tapped at equi-distant points, the taps 
being brought out to a dial switch. By 
the motion of this dial switch handle, 
more or less of the secondary winding 
could be thrown in series with the 
feeder, thus raising or lowering the 
pressure. A reversing switch was also 
provided by which the pressure of the 
regulating transformer could be op- 
posed to the bus pressure, if desired. 
This type is illustrated in Figs. 1 
and 2. 

Another type of regulator which 
was developed somewhat later is 
known as the "induction" type, and is 
illustrated in Figs. 3 and 4. 

In this regulator the variable volt- 
age of the secondary is secured by 
turning the movable core on which the 
secondary is wound to different posi- 
tions, thus linking more or less of 
the magnetic flux. If turned more 
than 180 degrees the secondary volt- 
age is reversible through its full range. 

This type is inferior in efficiency 
and power factor to the Stillwell type, 
owing to the presence of an air gap 
in the magnetic circuit, but its free- 
dom from sliding contacts renders it 
more suitable for use in cases where 
remote or automatic control is em- 

ployed. Remote control of regula- 
tors must be resorted to in situations 
where space is not available in the im- 
mediate vicinity of the switchboard, 
but may be had in a basement or on 
a gallery where it could not be utilized 
for other purposes so advantageously. 
Fig. 5 illustrates a typical equipment 
of this class, connections for which 
are shown in Fig. 6. An induction 
regulator is actuated by a small three- 
phase motor mounted on the regulator 
frame. A reversing switch, which is 
usually located on the feeder panel, 
enables the operator to move the regu- 
lator in either direction, thus raising 
or lowering the pressure. A limit 
switch is provided for the purpose of 
cutting the motor out when the regula- 
tor has been brought around to the 
position of maximum boost or choke. 
Hand control is also provided for use 
in emergency. 

Automatic feeder regulation has 
been attempted to a limited extent in 
conjunction with the use of motor- 
operated regulators. In the earlier 

fig. 4. 

forms, power was supplied by a motor- 
driven line shaft, from which a belt 
supplied power to each regulator pul- 
ley. The regulators were provided 
with direct-current magnetic clutches 
operated by balanced relays, which 
caused the clutches to engage the dial 
switch of the regulator and so raise 
or lower the pressure. 



March, 1908 

The adaptability of this arrangement 
was limited by the use of direct cur- 
rent, which is often not available in 
alternating-current stations. The use 
of a line shaft was also impossible in 

ing. A change in the current passing troublesome, and the added expense 

over the feeder causes the plunger to of the automatic equipment for each 

move and closes one of the voltmeter feeder makes it preferable in most 

contacts. This operates the relay cases to provide automatic control of 


On r- 
■ I 
1 \f 


I I 1 1 


i !lir 

CZ) i 


o- L -Vi"|i!rrr.oi 
— -< 1 1 ' ■ ii i , — ' 

t> | i i ; i i i i a 






FIG. 5. 

many places and the individual drive 
has superseded the earlier forms. 

The arrangement of automatic re- 
lay and motor-control circuits which 
is required is shown in Fig. 7. The 

?o reecter 



rriree-^7/7<x&e /o*v 


So Ger?era/br 

FIG. 7. 

switch, which in turn starts the motor, 
moves the regulator and raises or low- 
ers the feeder pressure. 

The series coil is made adjustable to 


Three phase Lotv7enj/or7 

Oeto// of l/m/t Str/tc/r 


solenoid of the contact making volt- 
meter contains a plunger which is 
actuated by an adjustable series wind- 
ing in opposition to a potential wind- 

permit its adaptation to feeders of 
various sizes and lengths. 

In practice the maintenance of re- 
lay contacts has been found somewhat 

bus pressure with hand- or remote- 
controlled feeder regulators. 

The load on lighting feeders usually 
does not vary rapidly, and automatic 
control of the bus pressure suffices 
to smooth out the small variations of 
pressure, so that the operator can 
easily care for variations due to grad- 
ual changes in the feeder load by hand 

The automatic regulator devised by 
Tirrell has proven very successful in 
the control of bus pressures. The 
general scheme of connections for this 
device is illustrated on Fig. 8, and the 
action may be described thus : 

The secondary circuits of the po- 
tential and current transformers of the 
generator are led through a solenoid 
in a compounding relation. The cur- 
rent section is subdivided so that dif- 
ferent rates of compounding may be 
secured. A movable plunger is actu- 
ated by this solenoid, which in turn 
actuates a counterweighted lever, the 
opposite end of which is equipped to 
make electrical contact in a relay 
circuit. The other contact terminal 
of this relay circuit is carried on a 
similar lever, which is actuated by the 
plunger of a direct-current solenoid. 
This solenoid receives current in pro- 
portion to the pressure at the exciter 
terminals. The relation of these con- 
tact-making levers is such that in- 
creased pressure at the exciter brushes 
tends to open the relay circuit, while 

March, 1908 



increased pressure at the main gen- 
erator terminals tends to close this re- 
lay circuit. The closing of the relay 
circuit demagnetizes the relay as one 
arm of the relay is continuously ex- 
cited in the opposite sense. As soon 


a portion of the exciter field rheo- 

Where there are several units in 
parallel in a station the regulator may 
be applied to the exciter for a part 
of them and the bus regulated for 


J ,,****/ earners 




*et.Ar coyr*CT3 







tO*mAf~**A rt»>4 

X5X § 

m c St**** to* 




FIG. 8. 

as the poles of the relay are demagnet- 
ized its armature is withdrawn by a 
spring. This closes a circuit which 
shunts the field rheostat of the exciter 
and greatly increases the terminal 
pressure at the exciter brushes. This 
increases the pull of the direct-current 
solenoid plunger and opens the relay 
circuit, thus weakening the pull. The 
result is a rapid vibratory action, 
which is kept up continuously at a con- 
stant rate while the load remains con- 
stant. As the load increases the cur- 
rent winding on the alternating-cur- 
rent solenoid exerts an increased pull 
on the plunger, which causes the lower 
contact of the relay circuit to move 
up toward the other contact and thus 
close the relay circuit sooner. This 
raises the exciter pressure and there- 
by the generator pressure until it has 
been restored to normal. The vibra- 
tory action continues as before, but 
the contacts are working in a slightly 
higher position in space, thus form- 
ing a "floating contact." 

A condenser is used to diminish the 
action of the arc at the contact which 
shunts the exciter rheostat. 

The ability of the shunt contacts 
to break the circuit is the limiting 
feature of the apparatus. This limit 
is reached at about 50 kw. on the ex- 
citer. Above this two or more breaks 
must be used in series, each shunting 

constant pressure, with the series coil 
of the alternating solenoid cut out. 
With this arrangement the bus pres- 
sure may be maintained constant at been worked out differently by two 

sure at each feeder end, without the 
use of pressure wires, has made the 
line-drop compensator an invaluable 
adjunct of alternating-current sys- 
tems. The function of the line-drop 
compensator is to introduce into the 
feeder voltmeter circuit a counter elec- 
tromotive force which shall reduce 
the reading of the voltmeter by an 
amount equivalent to the line drop, 
and therefore indicate to the station 
operator the pressure delivered at the 
feeder end. The compensator circuit 
is a miniature of the feeder itself, the 
pressure transformer representing the 
bus bar ; the compensator, the line ; and 
the voltmeter, the load. Since the 
feeder has both resistance and in- 
ductance the compensator must have 
two sections representing non-induct- 
ive and inductive drops. These sec- 
tions are subdivided into 10 or 20 di- 
visions, representing 1 per cent, or 2 
per cent, each, and equipped with dial 
switches, so that they can be adjusted 
to correspond with the drop in any 
feeder having a full-load drop of 1 
to 24 per cent. The counter electro- 
motive force is produced by passing 
current from the secondary of the 
feeder current transformer through 
the two sections of the compensator 
in series. 

The inductive, as well as non-in- 
ductive drop, being compensated for 
by the apparatus, the indications of 
the voltmeter are correct at all loads 
and at any power factor. 

The details of the compensator have 

100 Amp. 



< A ^ B 

/WVWWWWWW\| loooooooooooooootooooaqooooaoo I 


S (Www) 

To Load 


100 Amp. 

5 Amp. 



678910! 2 

® ® 





FIG. 9- 

any desired point by the insertion of 
an adjustable resistance in the pres- 
sure circuit of the alternating solenoid. 
The feasibility of having at the sta- 
tion an accurate indication of the pres- 

manufacturers in America. The ap- 
paratus as perfected by the originators 
of the device, the Westinghouse Elec- 
tric & Manufacturing Company, is il- 
lustrated diagrammatically in Fig. 9. 



March, 1908 

This company provides compensa- 
tors for a maximum drop of 6, 12, 24 
or 36 volts, the one illustrated in Fig. 
9 being a 24-volt compensator. 

The current from the secondary of 
the current transformer 6" passes 
through the inductive section B and 
the non-inductive section A in propor- 
tion to the load on the feeder. The 
ratio of the current transformer must 
be such that at its full-rated load the 

FIG. 10. 

current in the secondary will not ex- 
ceed 5 amperes. The inductive section 
is wound on an iron core, which serves 
also as the core of a pressure trans- 

The secondary winding is divided 
into four sections of five volts each and 
four sections of one volt each. The 
five-volt terminals are connected to the 
contacts numbered 1, 2, 3, 4, 5 and the 
one-volt terminals to the contacts num- 
bered 6, 7. 8, 9, 10. The arms may 
be independently adjusted, thus per- 
mitting any setting from 1 to 24 to be 
made, as in the following table : 


Per Cent. 




Per Cent 


















































The non-inductive section is similar- 
ly equipped and the settings are made 
in the same way. 

The pressure from the main pres- 
sure transformer C passes through 
the feeder voltmeter to terminal 6, 
through the two movable arms to 3, 
through the portion of the non-in- 
ductive section, which is included be- 
tween 3 and 5, thence through the in- 
ductive section by a similar path 

through the portions between 9 and 6, 
and between 4 and 5. It then returns 
to the pressure- transformer. In mak- 
ing this circuit the impressed pressure 
has been opposed by a counter electro- 
motive force of 10 volts in the non- 
inductive section and by 8 volts in the 
inductive section. 

The reading of the voltmeter is, 
therefore, reduced by the same amount 
as would be a voltmeter connected at 
the end of a feeder having a resistance 
drop of 10 volts (secondary) and a re- 
actance drop .of 8 volts at full load. 
If the normal secondary pressure de- 
livered to the feeder voltmeter is ap- 
proximately 100 these drops are also 
percentages of the secondary pressure, 
but if the secondary pressure on the 
voltmeter is no. or any other ap- 
preciably different voltage, the com- 
pensator figures cannot be considered 
as percentages of the secondary pres- 

The general external appearance of 
this type of compensator is illustrated 
in Fig. 10. 

The compensator, as worked out by 
the General Electric Company, is 
somewhat simpler in construction. 
The general scheme of connections is 
illustrated in Fig. n. 

In this type the current from the 
main current transformer is reduced 
in the ratio of 5 to 1 by a current 
transformer inside the case of the com- 
pensator. There is but one movable 
arm on each section and 8 points each 
of which represents 3 volts when 5 

eration of setting is easily accom- 
plished by any station attendant with- 
out danger of confusion or reference 
to a table of settings, corresponding 
to various percentages of compensa- 
tion. The general appearance of this 
type is shown in Fig. 12. 


With a feeder of No. o wire, 5000 
feet long, overhead wires 12 inches 

FIG. 12. 

apart, pressure 2200 volts at feeder 
end. frequency 60 cycles, current 
transformer rated 100 to 5 amperes. 
pressure transformer rated 2200 to no 
volts, how should the compensator be 
set ? 


la Load 


Too -Amp> 


J SAJi*>. 













amperes are following in the compen- 

The compensator in Fig. 1 1 is set 
so as to introduce in the voltmeter 
circuit an inductive counter electro- 
motive force of 9 volts and a non-in- 
ductive counter electromotive force of 
12 volts when the feeder is carrying 
full load. 

The points being numbered, the op- 

The full-load rating of the com- 
pensator being 5 amperes, that of 
the feeder is 100 amperes. The ohmic 
drop on a No. o feeder at 100 amperes 
is 0.2 volts per ampere per 1000 feet 
of two-wire circuit. Hence the ohmic 
drop is 100 x 5 x 0.2=100 volts or 
4.5 per cent. Likewise the inductive 
drop is 100 x 5 x 0.22=110 volts or 
5 per cent. 

March, 1908 



If the primary mains are designed 
to give not over 2 per cent, ohmic 
drop, the transformers 2 per cent, 
and secondary mains 2 per cent., the 
average ohmic drop from the feeder 
end to the consumer's premises should 
be about 3 per cent. ;the inductive drop 


premises. In this case the total ohmic 
drop is 4-5+3=7-5 per cent., while 
the inductive drop is 54-3=8 per 

If a Westinghouse 24-per cent, 
compensator were used, the setting of 
the resistance section would be 7^2 

.ompen&atoh CoA/A/rcT/OA/3 fott Tva Phase Umbalamccd Sx-stzm 

FIG. 13- 




FIG. 14- 

should be about 3 per cent. also. As- 
suming that these averages are ap- 
plicable to the major portion of the 
distributing mains, they may be added 
to the drop on the feeder, and the com- 
pensator set so that the drop on both ■ 
feeder and distributing system will be 
taken into account. The pressure may 
thus be regulated to give constant 
pressure at the average consumer's 

per cent, of 1 10 or 8 volts, and of the 
reactance section 8 per cent, of no 
or 9 volts. The resistance arms would 
therefore be set at 4-9 and the re- 
actance section 4-10. The operator 
will then keep the feeder voltmeter at 
no volts at all loads, this being main- 
tained as a standard pressure. 

With a General Electric compensa- 
tor having eight points on each part, 

the points have a value of three volts 
each, and the setting must be made on 
the nearest point. In this case the arm 
of each section would, therefore, be 
set at the third point. 

On a two-phase four-wire feeder 
the method of connection is similar to 
that used in the single-phase feeder, 
except that one equipment is required 
for each phase. The method of calcu- 
lating the setting for each phase is the 
same as in the case of a single-phase 
feeder. With a three-wire two-phase 
feeder, which always carries a bal- 
anced load, a compensator is required 
in the two phase wires only. But with 
unbalanced load one is required in 
each of the three wires. The connec- 
tions should be as shown in Fig. 13 
when the load is unbalanced. 

In calculating settings it must be 
borne in mind that the values of re- 
sistance and inductance per 1000 feet 
used in the case of the single-phase 
feeder are based on two wires, where- 
as in a three-wire feeder each com- 
pensator corrects the drop in one wire 
only. The values used for single- 
phase feeder resistance must, there- 
fore, be divided by two before being 
applied to a three-wire feeder, 
whether two phase or three phase. 

In case the common wire is equipped 
with a current transformer having a 
higher ratio than the other wires, this 
must be taken into account. Likewise 
if the common wire is larger than the 
other wires, the proper values must be 
used for this conductor. The allow- 
ance made for drop in the primary 
mains, transformers, secondaries, etc., 
should be added to the calculation for 
the phase wires of the feeder only as 
it is in phase with the drop in these 

With a two-phase feeder of three 
No. o wires similar in other respects 
to the single-phase feeder previously 
described, and with a current trans- 
former in the middle wire rated at 
150 to 5 amperes, the ohmic drop in 
the middle wire would be 5 x 150 x 
0.1=75 volts or 3.5 per cent., the in- 
ductive drop would be 5 x 150 x 0.11 
=82 volts or 4 per cent. The drop in 
the outer wires would be 5 x 100 x 0.1 
=50 ohmic and 0.55 inductive, or 
about 2.5 per cent. Adding the al- 
lowance of 3 per cent, for drop in the 
distributing mains, the compensator on 
the outer or phase wire should be set 
at 6 per cent, on each dial of the com- 
pensator. The compensator on the 
middle wire should be set at 4 per cent, 
on each dial. 

In the case of a three-wire three- 
phase feeder carrying unbalanced load, 
a compensator is required in each 
wire. For instance, if the feeder prev- 
iously used for illustration were a 



March, 1908 

three-wire three-phase feeder the 
ohmic drop in each wire will be 5 x 
100 x 0.1=50 volts, and the inductive 
drop 55 volts. These values are re- 
spectively 2.2 per cent, and 2.5 per 
cent, of the working pressure 2200 
volts. In this case the drop on each 
wire affects the pressure on two of the 
three phases. The compensators 
must, therefore, each interpose a 
counter electromotive force in the 
voltmeter circuits in proportion to the 
drop in the phase wire which it repre- 
sents. This drop must be expressed 
as a percentage of the working pres- 

The diagram of connections for this 
system is illustrated in Fig. 14. 

The allowance for drop in distribut- 
ing mains must be divided between 
any two compensators, as it is in phase 
with the working pressure. 1.5 per 
cent, should, therefore, be added to 
the 2.2 per cent, ohmic and 2.5 per 
cent, inductive drops, making the 
ohmic setting 3.7 per cent, and that 
of the inductive 4 per cent. 

In a three-phase four-wire system 
operating at 2200 volts between each 
phase and the neutral, the method of 
calculating the drop is as follows : 
With a feeder of four Xo. o wires 
running 5000 feet from the station as 
a three-phase feeder, the drop in each 
wire is 50 volts ohmic and 55 volts 
inductive. The working pressure be- 
ing 2200, this is 2.5 per cent. If the 
entire load of the feeder is delivered 
from this center of distribution the 
compensator on each phase wire 
should he set at 2.5 +3.0, or say, 6 

per cent, on each dial. That on the 
neutral should be set at 2 per cent, 
on each dial. If, however, the A- 
phase branches off with a neutral 
to a single-phase center of distribu- 
tion 2000 feet beyond, there must be 
added to the A-phase setting 100 x 
2 x 0.2=40 volts, or 2 per cent., mak- 
ing it 8 per cent, on each branch. If 
the other phases branch to similar 
centers of distribution at different 
distances, the drops must be figured as 
if they were single-phase feeders 
from the end of the three-phase trans- 


mission to the single-phase center of 
distribution. These drops must then 
be added to the three-phase drop 
above calculated. On four-wire 
feeders, which reach the limit of 
three-phase transmission within 3000 
feet of the station, it is usually un- 
necessary to install a compensator on 
the neutral wires, as the neutral drop 
is negligible, even with a consider- 
ably unbalanced load. 

The connections for a four-wire 
three-phase feeder are shown in Fig. 


fleuTXAi-j ^^ 

COA4 PUM-SATVA COA//V£CT/OAf£ foft foil* W/#£. TH&ZE f^M/tJE ^SXSTS M 

FIG. 15. 

Some Points in the Connecting and Repairing of 

Alternating-Current Motors. 


When a direct-current motor is 
spoken of as four-pole or six-pole, it 
is understood that the stationary, or 
field part, has that number of poles in 
it. These poles may be seen with the 
eye as large solid coils surrounding a 
metal core, but when a four-pole or 
six-pole alternating-current motor is 
examined no poles are visible. The 
field is then found to be an apparently 
continuous winding of wire, such as 
occurs upon the armature of a direct- 
current motor. The poles are there, 
nevertheless, and perform the same 
duty as in the direct-current machine. 
The reason they are not visible is be- 
cause the coils are wound flat in such 
a manner that the layers of any one 
pole overlap the windings of the ad- 

jacent poles, producing in the fin- 
ished machine the appearance of a 
continuous winding. If the connect- 
ing leads from the terminal block are 

"uiae-j ~LrS~ 

FIG. I. — two-phase; three-phase star, 
or y; three-phase delta connection. 

followed back, it will be found that in 
a single group perhaps five coils are 
joined together and connected to a 

similar group some distance beyond. 
In fact, if the leads are traced from 
the terminal block it will be seen that 
the seemingly continuous winding is 
cut up into many small groups. 

Each group of windings represents 
a collection of coils so connected by 
short stub connectors that the cur- 
rent passes through each coil of a 
group in the same direction. The 
coils of a group are, however, dis- 
tributed in successive slots. Each 
group represents a pole, every other 
one in a given phase being of one 
polarity and the next of opposite 
polarity. Thus we have north and 
south poles as in the direct-current 

The grouping of these coils is shown 

March, 1908 



diagramrnatically in Fig. I. It will 
be observed that in the two -phase mo- 
tor the two phases are separate. It 
is unusual to connect them. In the 
three-phase machine it is customary 
to connect one end of each phase to a 


common connection. The free ends 
are connected to the terminal block. 
This method of connection is known 
as the Y or star connection. Where 
all six terminals of the winding are 
connected to the terminal block we 
have the delta connection, the use of 
which is explained later. 

Fig. 1 shows the windings of a 
four-pole two-phase and of a four- 
pole three-phase machine connected in 
star and in delta. Six, eight, ten or 
more poles on each leg would simply 
include that many additional north 
and south poles. The reason for re- 
versing every other phase connection 
is to reverse the current through that 
pole and thereby produce a pole of 
opposite polarity from the one before 
it. Of course, north and south pole 
refers to a momentary condition only, 
as each reversal of current through a 
phase winding reverses the polarity of 
each pole. 

We will apply a Y three-phase 
winding to a stator (stationary part) 
having 24 slots, and therefore 24 coils, 
to be connected to each pole. The 
whole winding may therefore be 


paired off into groups of two coils in 
series. A start may be made from 
any point, as the position of a pole 
with reference to any mechanical line 
of the motor has no bearing at all. 
Fig. 2 A results in winding when 

each pole of one phase is hooked up, 
and Fig. 2 B when each phase is sim- 
ilarly connected. 

no-, 220-, 440-, 550- Volt Connec- 

Suppose each set of coils on a pole 
has 55 ohms resistance with four poles 
per phase, there would be a total of 
220 ohms per phase. Then, neglect- 
ing inductance, etc., for the sake of 
simplicity, 440 volts divided by 220 
volts equals two amperes. See Fig. 3. 
If 220 volts were applied to the mo- 
tor terminals, only one ampere would 
How. But if each two pairs of poles 
in series were paralleled with each 
other pair, as shown in the diagram, 
Fig. 5, then the same current would 
flow and the same torque would be 



~d6 tSJ 


FIG. 3 

Again, if all poles per phase are 
placed in parallel, as in Fig. 6, the re- 
sult would be the same. This is the 
plan upon which induction motors are 
wound, and as shown two amperes 
per coil will be obtained to give the 
required ampere turns for no, 220 or 
440 volts. It follows from the above 
that a motor wound, or put on the 


market as a machine for one voltage, 
may be changed over to a motor of 
similar speed and horse power, but for 
another voltage, by simply changing 
the loops to all series, series-parallel 
or parallel, as the case may call for. 
In motors of many poles, the number 
forming a group for any voltage will, 
of course, be one-half or one-fourth 
of the whole. For example, a 16-pole 
motor would have eight in series, or 
one leg for 220 volts, and four in se- 
ries for each no- volt leg. Five hun- 
dred and fifty volt motor windings are 
considered as "special" by manufac- 
turers, for no combination of groups 
will produce the right result for that 
voltage. A six-pole motor, or any 
machine having a winding which can- 
not be properly subdivided, must evi- 
dently have one particular special 
winding. If designed for 440 volts, 

with six poles in series, then three 
poles in series will give the 220-volt 
combination and the no-volt wind- 
ing would be special. Some manu- 
facturers, however, obtain the re- 
quired resistance for 220 volts by ar- 


ranging for the six poles to be in 
series at that voltage, in which case 
the three poles in parallel would give 
a no-volt winding and call for a spe- 
cial one for 440 volts. A repair or 
test man on getting a motor of this 
sort would have to follow the loops 
to determine which method had been 

"IfifiJ - W5 




FIG. 6. 

Some makers vary the method of 
connecting groups shown in Fig. I 
by connecting the north pole coils to- 
gether, and reversing through the 
south pole coils together, as shown 
for one phase of a winding at Fig. 7. 
This method of connecting is shown 
in several of the accompanying dia- 
grams and gives the same result as 

The reason for using the Y winding 
in three-phase work is evident. In 

3 t B, /? ft t 


the delta each leg has to withstand 
the full voltage of the terminals. In 
the Y we have a partial effect of two 
legs in series across the terminals, so 
that each phase has to lie insulated 



March, 1908 

only for -*?s of the line voltage, or 

practically 58 per cent. This mate- 
rially favors the insulation of motors 
with the windings connected in Y. 

This observation takes us into the 
usage of the delta winding. We have 
explained how pole groups are built 
on the unit plan. In a Y winding it 
is evident that phase A- A would not 
have 440 volts across its ends, but 58 
per cent., or about 255 volts. If a 
flow of two amperes were required, as 
in our example, then wire of lower 
resistance or slightly heavier cross 
section would have to be used than 
before. It often happens that at the 
end of a long transmission line, either 
220 or 440 volts, the standard voltages 
cannot be obtained, due to a heavy 
drop in the line or other effect. Sup- 
pose, for example, that in attempting 
to obtain 220 volts on the secondary 
only 180 volts or 250 volts could be 
secured. If one 220-volt motor were 
placed on a 180- volt line, the horse 

£5 Z X z &, H, 


power would be reduced in the ratio 
of (220) 2 to (180) 2 ; as nominal is 
to actual horse power. ( Horse power 
varies as square of voltage in alter- 
nating-current motors.) If a 20-h.p. 
motor were being considered, it would 
have its rating reduced to about 
l Z l A h-P- On the other hand, if this 
same motor were placed on the 250 
volts, its rating would change ap- 
proximately to 265*4 h.p. Neither re- 
sult would be desirable, for there 
would likely be too little power in one 
case and too much of a tendency in 
the other case to load up the motor 
with more work than its windings 
would stand. 

If, however, the motor were con- 
nected with all poles in series for 
440 volts and a delta connection used, 
the result would bring the windings 
to within a few volts of what would 
be required. It is customary to use 

a delta connection only in such a case 
as this. 

Most single-phase motors are in 
reality two-phase machines. The 
"running" winding will have many 
coils per pole and the "starting" 
winding few coils. In every other 
respect they are to be treated as two- 
phase motors. The General Electric 
motor has a straight three-phase 
winding, as was explained on page 25 
of the January issue of The Electric- 
al Age. In this motor the two ends 
of each phase are brought out of the 
motor, with the proper connections 
made in the starting box. The Wag- 
ner repulsion motor and other single- 
phase motors have their windings 
connected in accordance with the gen- 
eral plans described, and no trouble 
will be encountered if the leads from 
the terminal block are traced back in 
the winding. 

We have discussed heretofore only 
stator windings. Some types of 
motors have wound rotors or even 
armatures with commutators. The 
explanation given above, however, 
also explains wound rotors, as an in- 
spection with the eye will quickly 
show. Repulsion type motors with 
armatures having commutators are to 
be considered as it the armatures were 
for direct-current machines. 


The chief troubles of alternating- 
current motors are as follows : 

(i) Short-circuited coil. 

This will show itself quickly by 
heat, as the short-circuited section will 
act as the closed secondary of a 
transformer. The hand will readily 
detect the faulty spot or a sizzling of 
insulation if it continues very long. 

(2) Open Circuit. 

If in the stator of a two-phase 1 r 
Y-wound three-phase motor, the ma- 
chine will not start. If delta-wound 
it will start, but the torque will be 
reduced. If under the latter condi- 
tion the motor had to carry nearly a 
full load, it would not come to speed. 
Again, ammeter readings will show 
open circuits in the first two cases, and 
unbalanced readings on the delta. 

If in the rotor (wound rotor type) 
open circuits or unbalancing will 
show up on the ammeter. If all 
phases are open, of course the motor 
will not start at all. 

(3) Reversed Phase. 

If a part or the whole of one phase 
is reversed, the torque will be affected. 
If tested with an ammeter, unbalanced 
currents will be obtained when the 
motor is running. This trouble may 
also be located by tracing back the 
leads from the terminal block to find 

if the connections have been correctly 
made for the proper theoretical dia- 
gram. The rotor windings are simi- 
larly traced. 

< 4 ) Grounds. 

Open up all phases and test for 
grounds in accordance with the usual 
tests for such purposes. 

1 5 ) Humming. 

If there is a defective leg in any 
phase, so that full speed is not ob- 
tained, then the motor will hum quite 
loudly. Methods described above 
must then be used. If the motor at- 
tains full speed and hums, it prob- 
ably means loose laminations. The 
holding bolts must now be tightened 
or wedges driven in to take up space. 
In rare cases a few laminations may 
he jammed over with a cold chisel and 
hammer, in order to tighten up the 
core. This latter is only a method 
of last resort, as the insulation of the 
windings is apt to be injured by this 

If the rotor is off-center with re- 
■>pect to the stator, there will be mag- 
netic unbalancing and loud humming. 
This is easily remedied by turning up 
the machine so that the air gap is uni- 
form at all points. 


Motors from about one horse power 
to 300 h.p. are easily repaired away 
from the manufacturers' shop by the 
use of machine-made coils. Very 
small motors require many turns of 
small wire, and these are wound into 
the slots by hand. Such a winding. 
when completed, by being dipped and 
baked becomes so hard that it is al- 
most impossible to replace one coil 
without rewinding the whole motor. 
In very large motors the coils, when 
"set," become so hard that it is ex- 
tremely difficult to replace a defective 
one. In such cases a single coil may 
be dug out of the windings and the 

March, 1908 



a. W *, 

c a fi 






B z fl 2 /?, 3, 

"**A f, ft, 





3* /f z 3,/f, 

a o 99 

c a * 




two adjacent coils bridged across the duce the resistance of a given phase reduction of one or two per cent, will 
empty gap. This is common practice more than five per cent. If a parallel not unbalance enough to cause trouble 
where the coil removed will not re- or series-parallel winding is used, a of any kind, 

Investigation by the Public Service Commission 
of the Lighting Companies of New York 

THE inquiry of the Public Serv- 
ice Commission, First Dis- 
trict, into the affairs of the 
lighting companies of New York 
will cover the investigation into 
the franchises, property and opera- 
tions of the companies, inquiring 
into the methods employed by the 
companies, and each of them with re- 
spect to any discrimination in rates 
and whether such discrimination is 
undue, unreasonable, or unjust; 
whether contracts are required of cus- 
tomers as a condition to service, and if 
so, their nature, and whether legal, 
just and reasonable; emergency serv- 
ice and auxiliary or supplemental 
service ; regulations governing the in- 
troduction of wires upon the premises 
of customers and others, including the 
cost and charges therefor ; regulations 
governing the discontinuance of serv- 
ice and also the price charged for elec- 
tricity and any regulations governing 
the same ; the kind, condition and ac- 
curacy of meters used, the condition 
of the currents, wires, conduits and 
services, and generally the methods 
employed by the said corporations in 
generating and supplying electricity 
and in the transaction of their busi- 
ness ; and into the every matter and 
thing necessary or proper to inform 
the commission whether the property 
of a company is maintained and op- 
erated for the security and accommo- 
dation of the public, and in compliance 
with the provisions of law and their 
franchises and charters. 

The first subject taken up by the 
commission at the initial hearing. Feb- 
ruary 26th, was the matter of break- 
down service, to which the New York 
Edison Company objected, on the 
ground that the commission has no 
power or jurisdiction to go into this 
subject, in which protest several of 
the other companies also joined. 
Counsel for the commission read a 
letter from the New York Edison 
Company outlining a new service rate 
for "breakdown" installations. In its 
communication the company called at- 
tention to the fact that such a service 
is not a service in the usual sense, 
but insurance to protect customers 
against breakdown or overload and 
that it amounts to rendering the 
equivalent of a duplicate plant dupli- 
cated from the coal pile and boilers 
through the engines and generators to 
an independent generating and dis- 
tributing switchboard, with the equiva- 


lent of such a plant under constant 
steam pressure ready to serve at an 
instant's notice. The company calls 
attention to the fact that its engines 
and generators are always running 
to provide an adequate reserve, or 
their general service with other units 
revolving slowly in readiness to take 
up the load, all of which requires the 
operation of boilers, consuming steam, 
calling for many supply and main- 
tenance expenditures and the constant 
attendance of engineers and helpers. 

The tentative breakdown proposal is 
as follows : A service charge of $30 
annually for each kilowatt of installa- 
tion connected. 

In rating the installation each 16 
c-p. incandescent lamp should be taken 
as equivalent to 50 watts ; each arc 
lamp at 10 c-p. equivalents and each 
horse-power at 15 equivalents. 

Within the service charge of $30 
annually, customer may consume 
electric current at usual rates with- 
out additional charge. 

Service connections will be carried 
only to the building line, on which 
meters will be installed. Beyond this 
point connections must be provided by 
the consumer, who must furnish the 
throw-over switch. 

The first witness. Mr. Arthur Wil- 
liams, explained that the company did 
not distinguish between breakdown 
service and auxiliary service ; that 
there were only 123 such connections, 
which consumed during 1907 3^2 
million kw.-hrs. Most of these connec- 
tions were of long standing and none 
have been given within the last two 
years. Mr. Williams states that calls 
for this particular sort of service 
were very infrequent and not enough 
to attract any particular attention, un- 
til two years ago when the number 
increased and the company decided 
it would no longer supply such a 
service. The reason for discontinuing 
breakdown connections was given by 
witness as endangering the company's 
entire system by the possibility of hav- 
ing a very large installation thrown 
upon some part of the system not con- 
structed for the extra load. Mr. 
"Williams further stated that it would 
be absolutely necessary to provide ex- 
tra service distribution for the entire 
amount of breakdown connection, 
even with 2000 buildings connected, 
though he was not of the opinion that 
it would require the full generating 

In response to an inquiry as to how 
the company arrived at a service 
charge of $30 per kw. of breakdown 
installation, Mr. Williams stated that 
one point bearing on the price was 
the company's average return per kw. 
installed, which amounts to about $50 
a year. It was thought fair to charge 
three-fifths of this amount for break- 
down service. The $30, however, was 
not based upon any exact calculation 
as to cost : it represents about 10 
per cent, of $300, the investment 
cost for service. Mr. Williams' per- 
sonal opinion was that the amount 
should be 15 per cent, of $300, 15 
per cent, covering more closely the 
items of fixed cost. 

Mr. J. W. Lieb. Jr.. Third Vice- 
President of the company, was called 
at this point to explain what the $30 
represented and how the company ar- 
rived at this figure. He stated that 
the rate had been used and was, to 
some extent, current, it being the 
equivalent of the minimum guaran- 
tee that is sometimes exacted for the 
service of Si. 50 per year per 16 c-p. 
equivalent, which amounts to $30 per 
year. In response to an inquiryabout 
the S300 a kilowatt of capacity, Mr. 
Lieb did not have in mind any calcu- 
lations made on this line and would 
not undertake to explain Mr. Wil- 
liams' figures. He stated that they 
had made detailed calculations as to 
the proper stand-by costs. 

In response to a question whether 
$30 had been arrived at by taking 
three-fifths of the average of $50 per 
customer. Mr. Lieb stated : "That 
was a very important feature of the 
calculation, as I have already out- 
lined. That was one point of view, 
and one point of view leading to the 
S30 per cent, that I have already as- 
signed, and in considering the ques- 
tion from the other standpoint, from 
the standpoints and the considerations 
that I have indicated, again $30 
seemed to be a reasonable figure, and 
it was as the result of these con- 
siderations and general consultations 
that the figure of $30 was finally ar- 
rived at as a proper one to submit 
to the commission." 

In reponse to a question as to 
whether he wished to add anything 
further as to the way in. which this 
figure was attained, Mr. Lieb arr- 
swered. "I think, Mr. Commissioners, 
I have covered that subject fully." 

Mr. Hemmens, Counsel for the New 

March, 1908 



York Edison Company, at this point 
brought the witness to say that the 
$30 charge included the use of cur- 
rent up to that amount, so on this 
basis a breakdown customer was on 
the same basis as another customer. 

Here is a sample of some of the 
interesting testimony : 

(Q) You would consider then that 
possibly the same ratio would apply 
to breakdown service as to ordinary 

(A) Not necessarily, sir, possibly 

(Q) What do you estimate? 

(A) Well, I think that we had in 
mind that in an ordinary class of cus- 
sumer who would be likely to request 
a breakdown service which would 
mean, as a matter of fact, rather a 
wholesale consumer, that we should 
expect to be called upon for possibly 
between 60 and 75 per cent, of his 
connected installation. 

(Q) What is the percentage of the 
ordinary consumer? 

(A) About between 30 and 35. 

(Q) You consider it would be 
about double in the case of the 

(A) I think it would in this class 
of service. 

(Q) You say that is based on any 
data that you have? 

(A) On our best judgment. 

(Q) Do you know of any statistics 
that have been compiled to show ? 

(A) I have never seen any statis- 
tics as to what the demand in the case 
of breakdown service is, so I have no 
other guide than what I have indi- 

(Q) Do the facts regarding these 
123 customers — would they throw 
any light upon it? 

(A) I think not ; I think not. 

(Q) Why not? 

(A) Because they don't show. 

Mr. Weldon W. Freeman, Vice- 
President of the Edison Illuminating 
Company in Brooklyn, in response to 
inquiry, stated that he was connected 
with the company since 1889, first as 
assistant to the secretary and treas- 
urer until 1895, then assistant secre- 
tary until 1898; secretary until 1902, 
then secretary and treasurer for one 
year, and his present position since 

Mr. Freeman testified that their 
company exacted as a condition for 
breakdown service a minimum charge 
at the rate of $1 per year per 16-c-p. 
lamp installed, which is equivalent to 
$20 per kw. 

In early operations of the company, 
plants were connected without any 
such requirements as to guarantee, but 
after the company learned how un- 
profitable these connections were the 
guarantee was instituted some ten or 
twelve years ago. Only recently had 
they made inquiry into the actual cost 
when it developed for the year 1907 
each kw. connected installation cost 
$21.50 per kw. per annum. 

In making this estimate, all fixed 
charges of the company, except divi- 
dends and surplus profits, were in- 
cluded. Mr. Freeman stated that the 
maximum demand on their system was 
close to 50 per cent, of the total con- 
nected installations. He did not think 
it would be safe to figure any lower 
percentage for breakdown service than 
50 per cent. 

At the second hearing, March 5th, 
Mr. W. W. Freeman again took the 
stand for the Brooklyn Edison Com- 
pany and explained how their com- 
pany arrived at the $21.50 figure. He 

also explained that they had before 
omitted in their calculation the item 
of general expense, the inclusion of 
which would raise the stand-by charge 
from $21.50 to %2."j.2j per kw. He 
stated that the general expenses of 
their company were $370,643.26. 

Insurance $40,827.95 

Legal expenses and. . . 
damages 20,600 . 00 

Taxes 175,500.00 

Technical expenses.... 84,560.29 

Street and Installation 
maintenance and re- 
pair expenses 130,548.76 

Interest and discount 672,785.29 

Depreciation 242,931 .71 

With the total connection on the 
system at June 30, 1907, of 64,021 

Mr. Freeman stated that their aver- 
age receipts for the year 1907 were 
approximately $50 per kw. 

Mr. J. W.'Lieb, Jr., stated that in 
case breakdown connection assumed 
an importance which it has not hither- 
to had, that it would be necessary for 
the company to provide an automatic 
service switch so adjusted that an 
overload beyond that provided for in 
the contract could not be connected. An 
auxiliary device on the switch would 
give indication on the customer's 
switchboard when he reached within 
15 or 20 per cent, of his contract 
capacity, so that he would have con- 
siderable warning beforehand of the 
limit which he had asked for. It 
would also be necessary to provide 
an arrangement with the meter in 
order to avoid pumping back into the 
system and to avoid unregistering and 
the turning back of the meter. 

Study Men 


YOUR predecessors who have 
done their part as engineers 
in turning the forces of nature 
to the use of man have changed this 
world from one in which the winner 
was the man with the brute strength 
and physical bravery which gave him 
the power to win in a hand to hand 

By turning the forces of nature to 
the use of man, your predecessors as 
engineers have changed this into a 
world in which the winner is the man 
who thinks clearly, controls himself, 
and may be depended upon — the man 
who serves rather than the man who 

•From address delivered on Commencement Day. 
June 14, IQ07. Thomas S. Clarkson Memorial School 
of Technology, Potsdam, N. Y. 

Perhaps you think that I have ex- 
aggerated in crediting the engineer 
with all these changes. 

Think for a moment how the steam 
engine and other machines are the 
basis of your comfort. Think of the 
large part they have played in furnish- 
ing you the light and heat you have 
in your houses, the clothes you wear, 
the food you eat. 

The locomotive, the marine engine, 
the printing press and the telegraph, 
have made all the peoples of the world 
acquainted and changed them from 
enemies into friends. 

You, graduates, have been under 
the continuous influence of the teach- 
ers in school and college for sixteen 
to twenty years — for more than three- 

fourths of your life. You have ac- 
quired through their efforts. They 
have guided, encouraged and inspired 
you. To a large extent your knowl- 
edge has been selected by them and 
your views colored by them. You have 
learned from and through your teach- 
ers rather than from direct contact 
with facts. 

During this school and college 
period you have learned much from 
books rather than from teachers. But 
a book is simply the ideas of a man 
made visible and explained in the way 
which seems best to him. You seldom 
think of the man behind the book. 

If you prove to be a successful en- 
gineer you will pass through three 
periods with reference to the acquisi- 



March, 1908 

tion of knowledge and wisdom. First, 
the school and college period when 
you acquired through books and 
teachers. Second, the period com- 
prising the first ten or more years 
after you leave college, the period dur- 
ing which you will occupy subordinate 
positions and be in close contact with 
material facts. By that close contact 
with facts you will gain experience 
which will remedy, to a considerable 
extent, the inevitable defects of any 
education furnished by books and 
teachers alone. 

Just as rapidly and as certainly as 
you gain real success by showing abil- 
ity to make yourself useful in the 
world, and by using your ability, you 
will find your responsibility increased, 
the demands upon you increased, and 
will find that you cannot, if you are 
to accomplish most, remain in direct 
contact with all the facts of your daily 
work. You will enter into the third 
period with respect to the acquisition 
of knowledge and wisdom. You will 
find yourself in a position where you 
must acquire knowledge through your 
subordinates who are themselves in 
more direct contact with the facts. 
The chief engineer of a railroad, the 
chief engineer of a great government 
engineering bureau, like the Reclama- 
tion Service, the head of a great 
technical school, necessarily sees the 
facts of the work for which he is re- 
sponsible mainly through the eyes and 
brains of his subordinates. In the 
third, or executive, period then, as in 
the first, or school period, the success- 
ful engineer acquires knowledge and 
wisdom by utilizing the brains of 
other men. 

When you are in school and col- 
lege, you are. as a rule, learning things 
which were well known long before 
your time, you are acquiring knowl- 
edge which is well organized by the 
successive efforts of many men. teach- 
ers and authors. Because it is well 
organized knowledge, already worked 
over by many men, this concentrated 
experience comes to you from the 
past with comparatively little color- 
ing, due specifically to the last author 
and the last teacher in the series 
through which it passed to you. But 
it does come to you with high color- 
ing and in a distorted form, because 
the long series of authors and teach- 
ers have, as a rule, belonged to one 
profession — teaching — beacause they 
have all been thinkers, rather than 
doers. It is within your power, to a 
great extent, to remove the inevitable 
false coloring, and to round out the 
inevitably distorted form by heeding 
your own experience to be gained in 
the second period already referred to 
— the period during which you are to 
be in engineering in subordinate posi- 
tions in close contact with facts. 

But as you gradually, by being suc- 
cessful, pass into the third period in 
which you again depend upon utiliz- 
ing the brains of others, you will find 
that the facts you must deal with have 
not been known long, that they are 
not well organized, that they come to 
you through one man or through a 
short series of men only, and that, as 
a rule, the relations between the facts 
are but dimly perceived by the men 
from whom you get them. Under 
these conditions the facts and prin- 
ciples come to you highly colored and 
greatly distorted and but dimly out- 
lined because of the peculiarities of 
the man, or the few men, through 
whom you get them. It becomes, 
therefore, of prime importance to you 
to understand that man, or those men. 
To be entirely successful you must 
study men. 

An engineer does very little directly 
without the intervention of other men 
between him and his accomplishment, 
even when he is in minor, subordinate 
positions. Even the levelman is de- 
pendent on his rodman and recorder. 
The inspector on construction may see 
with his own eyes, but he produces 
changes only by operating through a 
foreman or perhaps a chain of several 
men, including the engineer to whom 
he reports, the contractor, the con- 
tractor's foreman, and finally the 
workmen. The draftsman may seem 
to be directly in contact with his work, 
but he really accomplishes something 
only as he succeeds by means of draw- 
ings in guiding the skilled workman 
whom perhaps he never sees. In each 
of even these simple cases the ef- 
fectiveness of the engineer is condi- 
tioned in part on his accurate under- 
standing of the thoughts and feelings 
of the men through whom he works. 

As an engineer rises higher in the 
organization with which he works, his 
field of influence becomes larger, but 
the line of men through whom he 
works to produce material results also 
lengthens. He works to an increasing 
degree through other men and it is of 
increasing importance that he under- 
stands other men. Or, if he fails to 
know men he is apt to fail to rise. 

If you are to succeed — to be valu- 
able in the world — to know is not 
enough, you must make others to 

As soon as you are well started in 
studying men you will find yourself 
studying the need and purpose of or- 
ganization. For as soon as you fully 
realize what great differences there are 
in their principal characteristics, and 
even how widely the capabilities of a 
given man may van at different stages 
of his life, you will realize why and 
how it is that a group of men working 
together as an organization may ac- 
complish much more than the same 

men could if they worked inde- 
pendently, as individuals. 

A very common conception of or- 
ganization is that it is an arbitrary 
arrangement by which orders are 
transmitted by various steps, through 
different groups of officials, from the 
man at the head of the organization 
to the many men who form the rank 
and file and do the actual work. Many 
graduates have shown that they be- 
lieve that the way for a man in a high 
position to get a thing done is to order 
it done. Poor and inefficient admin- 
istrators may do it that way. The 
successful administrators are men who 
act on the principle that their business 
is to administer unto those below 
them in the organization in three 
ways. First, by putting them into 
such places and under such conditions 
that they can do their best; second, 
by giving them orders necessary to 
show what is expected of them ; and, 
third, by enlisting their wills, as well 
as their bodies and minds, in the work 
of the organization so that they will 
do their best. The first and third of 
these, the average graduate has never 
seriously thought of. He sees in the 
administrative officer the man who 
orders. The successful administrator 
finds his time so thoroughly filled with 
the first and third kinds of administra- 
tion, with putting each man in the 
place and under the conditions most 
favorable to his effectiveness, and 
with enlisting in the service the will 
of the man, that orders fill but a small 
part of his horizon. 

The men near the top in an organ- 
ization normally do the most difficult 
work. Normally, they are the men 
who work most intensely and for the 
longest hours. In the great organiza- 
tion with which I am connected, the 
civil service of the United States, this 
is so commonly recognized that it calls 
forth no comment to see the rank and 
file leave at four-thirty and come at 
exactly nine, while others who are in 
responsible control of the organization 
work early, late and strenuously. 

To attain to the highest success as 
an engineer you must not be the type 
of man who knows how to do things 
excellently, but cannot tell others how 
to do them — the man who gets knowl- 
edge abundantly, but can apply it only 
through his own fingers. Instead of 
devoting your energy simply to in- 
creasing you own output by fifty or 
even one hundred per cent., it is far 
better — you make yourself more use- 
ful to the world — by using your en- 
ergy to increase the output of each 
of one hundred men by ten per cent. 
The world recognizes this by award- 
ing the prizes to the administrators. 

Some Points About Series Transformers 


IT is commonly known, but not so 
commonly appreciated as it should 
be. that the secondaries of series 
transformers should be always either 
short circuited when there is current 
flowing in the primaries, or else con- 
nected so that the current delivered 
may not be constrained from its nor- 
mal value or phase by the reaction 
of other series transformers. The 
bad effects of such constraint of the 
secondary flow of current are as fol- 
lows : 

A serious voltage rise occurs upon 
the whole series system fed by the 
transformers. This often results in 
someone getting a bad bump who 
touches the series wiring, and has in 
at least one instance resulted fatally. 
Although a voltmeter placed across 
the open secondary of the series 
transformer that produced the above 
fatality showed only 400 volts, yet a 
consideration of the case will show 
that while the "root mean square" 
voltage as shown on the meter was 
400 the instantaneous voltage waves 
were probably over 1500 volts. 

The voltage rise on opening the 
secondary series connections has in 
several instances been observed to 
break down insulation in instruments 
connected to series transformers and 
has been seen to strike an arc over 
one thirty-second inch and to draw an 
arc of over an inch when full load 
was on the transformers. 

Instruments are given in the fac- 
tory higher insulation tests on shunt 
circuits than on series circuits, but in 
practice more than otherwise are the 
series circuits given the greater in- 
sulation strain due to open circuited 
series connections. 

The iron of series transformers is 
apt to run hot enough to destroy the 
insulation if the secondary is left open 
for a length of time with load on the 
primary. This is on account of the 
excessive saturation of the iron clue 
to the prevention of the secondary 
ampere turns from neutralizing the 
effect of the primary ampere turns in 
magnetizing the iron. 

The reasons for the excessivelv 
peaked voltage wave delivered by the 
open secondary of a heavily loaded 
series transformer are as follows : 

When the secondary current is not 
allowed to flow and neutralize almost 
entirely the magnetizing effect of the 
primary current, then the whole full 

load value of the primary current 
goes to magnetize the iron and the 
result is that the iron is strongly sat- 
urated. This saturation obtains over 
most of the time occupied by a half 
cycle, and in a very short time when 
the primary current is passing 
through its zero value, the whole flux 
reverses and grows to saturation in 
the opposite direction. During this 
short time the voltage wave climbs to 
its high peak and dies away again. 
This is the interval in which the man 
gets his "bump" and the instrument 
gets its insulation strain. A peaked 
wave shows only a low voltmeter 
reading compared to its maximum in- 
stantaneous value, and a voltmeter 
connected to the open secondary al- 
lows a little current to flow, thus re- 
ceiving less than the true open circuit 
voltage of the series transformer. 

Some cases of voltage rise have 
been observed as follows : 

First : The secondary of two series 
transformers were erroneously con- 
nected in series and fed some meter 
circuits. The primaries being on dif- 
ferent phases of the power circuit. 
Each transformer tried to deliver cur- 
rent in phase with its primary current, 
but both secondaries being in series, 
constrained the current to be alike in 
both coils. The result was, some- 
what hot iron and a voltage rise. 

Second : On a three-phase line 
there was a series transformer pri- 
mary in each wire of line. The 
secondaries were connected in star, 
but one secondary was reversed. Each 
secondary feeds an ammeter, and the 
circuits after passing through the 
ammeters came together in a star. 
The instantaneous value of the cur- 
rent which the reversed transformer 
tried to deliver toward its ammeter 
w as opposed by the resultant of the 
instantaneous currents fed from the 
other two transformers. The result 
was that two of the ammeters showed 
an incorrect reading, and the third 
showed zero. There was a voltage 
difference generated between the star 
point of the transformers and the star 
point of the meters. Upon connect- 
ing these two star points by a wire 
the three meters read correctly, and 
the flow of current in this star con- 
nector was twice that in one of the 

Third: When the reversed trans- 
former in the above case was cor- 

rected there Was no flow of current in 
the star connector and it was re- 
moved. However, when a partial 
ground would occur on the line a 
spitting was reported to be heard in 
the meters. This can be explained by 
a voltage rise between the star points 
being generated due to the fact that 
the instantaneous algebraic sum of 
the primary currents in the series 
transformers did not add up to zero 
and hence the secondary currents 
tried to add up to a value other than 
zero, but on account of there being 
no star connector this excess current 
could not flow and hence the voltage 
rise was produced. 

Centrifugal Pump Trouble 


The recent experience of a manu- 
facturer of centrifugal pumps with 
outfits consisting of a centrifugal 
pump connected to a shunt- wound, 
direct-current motor manufactured 
into a direct-connected unit, may be 
of interest as showing some of the 
peculiar phases of electric drive. The 
head against which the pump was to 
work called for a speed in the pump 
of about 1 180 rev. per min. As the 
competition was severe the pump 
manufacturer purchased one of the 
cheaper motors in the market of 1200 
rev. per min. The motor was sent to 
the pump shop for mounting, and the 
set having been tested was shipped 
to the purchaser. 

The outfit was duly installed and 
connected up and current finally 
turned on. The attendant found 
that the pump did not pick up its 
load. He was sufficiently familiar 
with direct-current motors to know 
that by shifting his brushes a bit the 
speed could be raised above normal. 
Tic shifted them until the pump 
caught its load and started in to work 
successfully. Finding everything 
working satisfactorily he went about 
his duties in another part of the build- 
ing but returned in one-half hour 
just in time to see his fuses blow out. 
but not before he caught by the 
sound that the set had been racing 
at unknown speed. He put in new 
fuses and jiggered the brushes back 
to the 1200 rev. per min. point, hut 
as the indicator showed that the tank- 
was full, he left the set without start- 




March, 1908 

ing it again. In twenty minutes he 
returned to find that more water was 
needed in the tank and accordingly 
started up the set. The motor now 
picked up its load at once although it 
had refused previously to do so at 
this point. He let it run a while, and 
making sure that everything was 
right, went away for about fifteen 
minutes. On returning he again 
found the pump racing and so he now 
moved the brushes down to about the 
normal noo rev. per rriin. point, when 
the speed, according to his ear, be- 
came all right again. After a five- 
minute run at this speed the tank be- 
came filled and he pulled the switch. 
In less than a quarter of an hour he 
received a call from the office to start 
the pump on account of low water, 
and forgetting where he had left his 
brushes started up on the noo rev. 
per min. speed. The motor picked 
up its load at once and continued 
without any trouble. The following 
morning when he tried to start on 
the noo rev. per min. point the motor 
would not pick up its load nor would 
it start on the 1200 rev. per min. 
point ; it was only when he shoved 
the brushes to the 1300 rev. per min. 
point that he could start the pump. 
Intermittently all day long he would 
sometimes be able to pick up load on 
one speed or another. 

This peculiar condition when re- 
ported to the pump manufacturer 
caused considerable speculation as to 
the cause, as the set had been tested 
before it left the shop. The trouble 
had every indication of being an elec- 
trical one and the motor manufac- 
turer was called into consultation. 
He very quickly located the trouble 
and corrected it. In the first place, 
the set when first put up on their test 
floor was put to work against an 
artificial head supposed to be equiva- 
lent to the service conditions but 
probably not equal to it. When in- 
stalled at the customer's work and 
first started the motor was cold, and 
1200 rev. per min. would not quite 
pump water to the tank. The at- 
tendant having raised the speed to 
about 1300 rev. per" min. obtained 
speed enough and delivery started. 
The motor carrying its full load 
quickly warmed up and the increased 
heat of the field-winding soon caused 
a higher resistance in the fields. This 
higher resistance weakened the field 
and increased the speed of the motor 
slightly. This change in speed 
caused the pump to increase its load 
as the square of the speed change, so 
that the slight excess of motor speed 
produced a relatively heavy pump 
overload. The motor now began to 
draw still more current followed by 
more heating and still higher speed, 
following N this cycle until ' the fuses 

finally blew out as the attendant first 
came into the room. With the 
brushes shifted back to the 1200 point, 
the set was started up the second 
time with the fields still hot, so that 
1300 rev. per min. were obtained in- 
stead of 1200 rev. per min. The 
brushes after some time were shifted 
back to the 1100 rev. per min. point, 
where the load was thrown off. The 
set was started up the third time 
while the fields were still hot enough 
to give 1300 rev. per min. from the 
1 100 point. Of course the following 
morning the motor would not pick up 
its load at either the 1100 or 1200 
points because the motor had cooled 
down over night. 

A motor wound for 1300 rev. per 
min. with a compound winding ad- 
justed to keep the speed change 
within two per cent, was substituted 
for the shunt motor and no further 
trouble resulted. Ordinarily a shunt- 
wound motor would not have changed 
in speed as much as this one did, but 
the motor manufacturer had put into 
his yokes an inferior iron, with the 
result that there was practically no 
regulation in the motor. 

Commercial Day Program N. E,. 
L. A. Convention, 1908 

Thursday. May 21st. has been set 
aside by President Farrand as com- 
mercial day for the annual convention 
of the National Electric Light Asso- 

The committee, composed of C. W. 
Lee, J. Robert Crouse, John F. Gil- 
christ, George Williams, Howard K. 
Mohr and Frank B. Rae, Jr., has been 
busily engaged in preparing the pro- 
gram for the day. It is the purpose 
of the committee to have the papers 
which will be presented consume the 
least possible time : the major portion 
of the two sessions being devoted to 
discussion. The following is the pro- 
gram as outlined by the committee : 

( 1 ) Special Feature : 

Relationship between the en- 
gineering and commercial 
departments by a prominent 
electrical engineer. 

(2) Preparation for a Campaign : 

(a) Field work and other es- 

(b) Analysis of customers' 

(c) Proportion of 1 a m p 
equivalents lost to lamps 
connected — showing per- 
centage in cities of varied 

(d) Policy of handling com- 

(e) Policy of handling col- 

( 3 ) The Contract Agent and the 
Representative : 

(a) The contract agent — his 

(b) The district representa- 
tive — his possibilities, 

(c) The special representa- 

1. The sign expert, 

2. The power expert, 

3. The woman represent- 

(d) Solicitors meetings — their 

(4) The Display Room: 

(a) Appointments and meth- 

(b) Value of special demon- 

(c) Value of electrical and 
food show exhibits. 

(5) Advertising: 

(a) What is being done? 

(b) Why? 

(c) Results. 

(6) Publicity: 

(a) Methods to create proper 
public sentiment, 

(b) Dormant publicity oppor- 
tunities of lighting com- 

( 7 ) Creating Demands for Electric- 

(a) The creative principle, 

(b) Xotable examples, 

(c) Stereopticon talk upon 
outlines and sign lighting — 
showing progress in large 
and small cities. 

(8) Evolution of New Business 

(a) Examples of central sta- 
tions that have continued 
methods during depression, 

(b) Strong plea for up-keep 
of commercial departments 
and advertising, 

(c) Opportunities for creat- 
ing business along existing 

(9) The Electrical Contractor : 
Symposium : 

(a) What he is doing to as- 
sist in creating greater de- 
mands for electricity. 

(b) Specific examples. 

( 10) Co-operative Commercialism : 

J. Robert Crouse. 

(11) Illuminating Engineering as a 
Commercial Factor : 


By V. R. Lansingh. 

(12) Report of Committee on Solici- 
tors' hand-book: 

Award of prizes offered by 
Co-operative Electrical De- 
velopment Association. 

March, 1908 



Questions and Answers 

Question. — / have to wire up a 
three-horse power, alternating-current 
motor which is not provided with any 
compensating starting device. Nozv 
the full load amperes of this motor 
are 10 per leg. The starting am- 
peres, according to the bulletin, will be 
about 40. If I fuse my line switch for 
10 amperes they will blow out every 
time I start. If I put in 40-ampere 
fuses I will have no protection against 
overload zvhen the motor settles dozvn 
to zcork. 

Answer. — Wire according to the 
sketch herewith shown. 
















40 Amp. 

15 Amp. 

By this method the 15-ampere fuses 
will be cut out during starting, but 
cut into service when the motor has 
reached load speed and current is 
down to its running value. 

Question. — Does not a slow-speed 
motor take less current through the 
meter than a high-speed one, because 
its speed is not so great and there- 
fore would not require so much cur- 
rent to drive it as a high-speed motor ? 

Answer — Both motors will take the 
same current through the meters. 
A motor does a certain amount of 
horse-power work and draws that 
amount of horse power of electrical 
energy through the meter. The unit 
of horse power is made up of two 
things, a pull measured in pounds and 
a distance through which that pull is 
exerted, measured in feet. A multi- 
plication of these two gives foot- 
pounds of energy (one pound by one 
foot equals one foot-pound). If one 
motor be built with a pulling effect 
at the surface of its pulley equal to 10 
lb., and we revolve this pulley through 
a distance of 3300 ft. a minute (ap- 
proximately 1 100 rev. per min.) we 
have a i-h.p. motor. If, however, 
we had taken a motor having 

30 lb. at the pulley and had re- 
volved it through only 1 100 ft. (or 
about 375 rev. per min.) we would 
still have a i-h.p. motor. In both 
cases the foot-pounds were the same, 
and as the customer pays for the elec- 
trical equivalent of the foot-pounds 
(watts) his bills would be alike in 
both cases. 

Question. — / have a number of belt 
drives to install where my changes in 
speeds will be all the way from 3-1 to 
10-1. What is the limit of change be- 
tween any tzi'o pulleys according to 
best belt-drive practice before there 
z\.'ould begin to be too much slippage 
on the sm-all pulley end due to the 
belt not having enough contact sur- 

Answer. — Millwrights seldom in- 
stall over a 6:1 reduction between pul- 
leys unless the distance between 
centers is very great, that is, consid- 
erably over 15 ft. If obliged to make 
a 10- 1 reduction you should belt 
motors to a countershaft and belt to 
drive shaft again, making two reduc- 
tions. As an alternative you can 
purchase a motor carry an idler 
pulley which will keep the belt lapped 
over about two-thirds of the motor 
pulley. By using this drive you can 
belt directly from shaft to motor pul- 
ley on even a 10-1 reduction. 

Question. — We are going to install 
a centrifugal pump driven by a direct- 
current motor. Should we purchase 
shunt or compound-wound machine. 7 

Answer. — Buy a compound-wound 
motor so that t!._ speed of the shaft 
may be as nearly as possible constant 
under all conditions. Read the article 
on another page telling of a peculiar 
pump trouble. 

Question. — Will you please inform 
me through the valuable columns of 
your Journal the formula for finding 
the safe carrying capacity in amperes 
of rubber-covered zvire? 

Answer. — The formula is usually 
given as 

d i 

I = , 

where I is current in amperes and d is 
diameter in mils, or thousands of an 
inch. The values are given in the 
table : 

B. * S. GAUGE 









10- 9 


Strand . . 

8- 7 





















Electrolytic Copper Refining 

A current of electricity passed be- 
tween two electrodes of copper sus- 
pended in a solution of copper sul- 
phate will dissolve copper from one 
electrode, carry it to the other elec- 
trode and deposit it thereon. This is 
the process by which about 400,000 
tons of copper, more than one-half 
the raw copper production of the 
world, is refined. There are about a 
dozen electrolytic copper refineries in 
the United States, and these supply 
over 86 per cent, of the world's out- 
put of electrolytic copper. 

Copper may be refined in less ex- 
pensive ways where the raw copper 
does not contain gold and silver, but 
where these metals exist in the raw 
copper the electrolytic process is justi- 
fied by the value of the by-products. 
Ulke has estimated that over 27,000,- 
000 oz. of silver and 346,020 oz. of 
gold are recovered annually from the 
slimes of the American (U. S.) cop- 
per refineries. We are therefore 
largely indebted to the presence of the 
precious metals in copper ores for the 
pure copper so much used in the elec- 
trical industries. 

The anodes of cast raw copper are 
prepared in casting machines, a mod- 
ern example of which is the revolving 
furnace at the Tacoma refinery. This 
is constructed on the model of the 
black-ash revolving furnace used in 
the Le Blanc alkali industry, the 
anodes being cast by pouring the 
molten metal on a series of movable 
trays and molds. These anodes are 
cast with lugs, by which they are sup- 
ported upon two bus bars running 
along each side of the vat. 

The cathodes in the multiple proc- 
ess described below are thin sheets of 
pure copper obtained by electro- 
deposition in special vats. They are 
provided with devices to prevent 
buckling and are supported by hooks 
from rods running transversely across 
the vats. 

The vats are of wood, lined with 
lead and supported on insulators, and 
therefore resemble the tanks used for 
station storage batteries. The num- 
ber of electrodes in each vat is some- 
times as great as 60 and the weight 
of anode metal per vat may now 
amount to 6^4 tons. 

The anodes and cathodes in each 
vat are generally connected respect- 
ively in parallel and the vats in series. 
This is known as the multiple or 
Thofern system, but another one 
known as the series or Hayden sys- 
tem is coming into use. In this system 
only the two end electrodes in each 
vat are connected to the current leads, 
and all the intermediate ones are sup- 
ported on insulators. The current 
passes from one end of ' the vat 



March, 1908 

through the electrolyte and through 
the secondary electrodes, and while 
copper is dissolved from one face of 
these, it is deposited on the other face. 
The cathode faces have to be specially 
prepared in order that the deposited 
copper may be stripped off. This 
process eliminates the use of special 
cathode sheets and separates the elec- 
trodes of opposite voltage so as to 
reduce the risk of short circuits. 

The current density varies from 13 
to 40 amperes per square foot in dif- 
ferent refineries. A high current 
density involves extra operating risks, 
but is desirable in order to obtain a 
quick turn-over of the copper. 

The latest practice is to use high 
speed homopolar generators with taps 
for 125, 250, 375 and 500 volts, the 
voltage used depending on the num- 
ber of vats to be connected. These 
generators offer the important ad- 
vantage of having no commutator. 

The charging and discharging of 
the vats are effected by traveling 

The precious metals are not dis- 
solved appreciably by the sulphuric 
acid radicle and therefore fall to the 
bottom of the vats as slime. A slight 
amount of silver which may go into 
solution is precipitated by hydro- 
chloric acid in the electrolyte. The 
slime is removed from the vats 

periodically and boiled in sulphuric 
acid to remove traces of copper and 
is separated from the copper sulphate 
by filtration. The mud is thoroughly 
washed, pressed into cakes, and the 
gold and silver extracted from the 
dried cake by metallurgical and chem- 
ical parting methods. 

Electricity In Construction Worh 

Steel sheet piling has made a place 
for itself in construction work owing 
to its superiority over wooden sheet 
piling, but it has a very serious dis- 
advantage, in that whenever it is 
necessary to cut such piling off at a 
desired level it is a very expensive 
piece of work to do by hand, even 
when the assistance of power-operated 
drills and hammers is available. In 
the construction of the foundations 
for an extension of the Hoffman 
House in New York, interlocking 
channel bar steel sheet piling was 
used, the sectional area of metal per 
lineal foot of piling being 14 sq. in. 
In cutting this off to level, using elec- 
tric drills and cold chisels, the labor 
cost $9 and the electricity $13. ma- 
king a total cost of §22 per lineal foot. 
This was so expensive that it was 
decided to endeavor to burn the piling 
off by the use of an electric arc. 
A connection was made to the lines 

of the United Electric Light and 
Power Company, delivering single- 
phase 2500-volt current and four 20- 
k\v. transformers were installed, con- 
nected in multiple to deliver current 
at 50 volts potential. One side of the 
circuit was grounded on the steel 
piles while the other was connected 
to a carbon electrode mounted on a 
long wooden handle. The electrode 
consisted of a carbon bar 12 in. long 
and \ x /\ by Y\ in. clamped between 
two copper plates. A horizontal bar 
with a sliding suspender was pro- 
vided to guide the electrode. The 
man operating this device was pro- 
tected by an asbestos mask, large 
black goggles and gloves, so that no 
portion of his skin was exposed to the 
glare of the arc or its heat. 

The operation of the device was 
very simple, it being only necessary 
to make a contact with the steel pile 
and form an arc, care being used to 
avoid breaking the arc after its for- 
mation. The arc consumed about 
650 amperes at 50 volts potential and 
the cost of cutting off the piles by 
this method was $1 per lineal foot. 
The foregoing data was supplied by 
the Thompson-Starrett Company, 
contractors for the building, and is of 
interest as illustrating the adaptabil- 
ity of electricity to one of the diffi- 
cult problems met by contractors. 

Downward Illumination 

THOSE who have been long in 
the lamp business will re- 
member the familiar dark spot 
under the old Edison hair-pin fila- 
ment. Actual test showed the down- 
ward distribution of this form of 
carbon filament to be about 2 c-p. 
on a 16-c-p. lamp. While the 
dim spot on the floor or table be- 
neath the lamp was of course less 
objectionable than the broad black 
shadow of the gas fixture which 
it replaced, nevertheless the down- 
ward distribution of the filament 
was quite unsatisfactory. Some 
years later the idea was introduced of 
a double hair-pin filament, each part 
having about half the length of the 
first filament. This was the original 
Bryan Marsh double filament lamp. 
It gave 4 c-p. in downward intensity 
on a 16 c-p. lamp and was satis- 
factory progress. Following closely 
upon this development came the oval- 
anchored type of filament adopted by 
the General Electric Company which 
is the universal type of Edison lamp 
now in service. A 16-c-p. lamp using 
this filament shows 7.2 c-p. vertically 
downward. The light distribution 
curves corresponding to these various 
figures are shown in the accompany- 

ing cut. The particular set of tests 
from which these curves and figures 
are taken was made in 1904 by the 
United States Bureau of Standards, 
Washington, D. C, under the super- 
vision of S. W. Stratton, director of 
the Bureau. 

The deficiency of the oval-anchored 
filament in downward illumination 
has, however, been sufficient to pro- 
duce several types of lamps which 
surpass it in this respect. 

The ordinary incandescent lamp is 
usually placed some feet above the 
point where light is desired ; and ex- 
cept in the general illumination of 
public buildings, the strongest light is 
needed directly under the lamp or in 
the lower quarter circle of its distribu- 
tion. Illustrations of this may be 
found in the lighting of machines, 
work benches, desks and counters. 
The use of a reflector would be un- 
necessary if the filament were ar- 
ranged so as to have the largest 
amount of possible light in a general 
downward direction instead of in a 
sidewise direction. To appreciate the 
force of this point it is only necessary 
to hold an Edison lamp horizontally 
and then vertically, noting the great 
increase of light in the former posi- 

tion. As a result of continued thought 
on this problem we have the double 
round coil used in the Columbia lam]). 
the Tipless Lamp Company's lamp 
and the Sunbeam reflector lamp. This 
filament shows 9.8 c-p. downward in 
a 16-c-p. lamp. Lastly come the 
Shelby double flattened coil filament 
showing it. 3 c-p. downward, the 
Sterling spiral showing 15.3 c-p., and 
the Wormley or so-called downward 
light lamp, showing a maximum in- 
tensity of 16 c-p. downward. In the 
downward light uniformity of hori- 
zontal distribution is somewhat sacri- 
ficed : the maximum horizontal in- 
tensity of light in this lamp being 
twice as great as the minimum. In 
most of the other types, the horizontal 
distribution is nearly uniform. Com- 
paring all the lamps, however, on the 
basis of equal mean spherical candle- 
power, that is for a given total flux 
of light, the Wormley lamp shows 
much the greatest intensity of light in 
the quarter circle downward. In this 
respect the Sterling spiral filament is 
a close second. 

Both of these types of lamps are 
rapidly gaining favor in sign work- 
where the maximum light is needed 
through the end of the bulb. In this 

March, 1908 





\ S 




?/ r 


• — 


•VVv \ 

\ v^ Jsr-~~ — 

/ * 

1 * 

1*9 / 

Z'nr, OV3I 





' /.*/- 







/V»k /# 






\ M 












/* 1 













s ,. 

x ; 







N \ 









<fef Coll 








class of work a 2 c-p. downward suc- 
cessfully displaces a 4 c-p. oval-an- 
chored filament lamp, having as much 
outward illumination with the same 
current consumption. 

There is no commercial reason why 
one of the latter types should not dis- 
place the oval-anchored type entirely 
for car lighting. In cars where lights 
are placed vertically over the seats 
without reflectors, as in the Inter- 
borough and New York Central cars, 
a 16 c-p. downward light would 
throw twice as much light on the pas- 
sengers' newspapers as a 16 c-p. 
Edison lamp ; or on the other hand an 
8 c-p. downward would give as much 
as a 16 c-p. Edison. With the Ster- 
ling or Shelby light the advantage 

would be almost as great. In store 
lighting, too, the advantage of this 
class of lamps is evident not only in 
throwing light down upon the show- 
case and counter, but also diagonally 
upward upon shelves of goods. It is 
plain from a study of the distribution 
curves that such a lamp hung verti- 
cally over a counter will illuminate 
the shelves of goods displayed from 
ceiling to floor in a more nearly uni- 
form manner than the oval-anchored 
filament. Rays of light proceeding 
in a horizontal direction travel the 
shortest distance to illuminate the ob- 
jects upon which they fall. The in- 
tensity of the horizontal light from 
these lamps is less than that which is 
thrown in a diagonal direction and 

which has to travel a greater distance 
to illuminate its field. In using an 
oval-anchored filament the most bril- 
liant rays are given out sideways, and 
as its horizontal illumination has the 
shortest distance to travel, objects ap- 
proximately in a horizontal plane are 
brilliantly illuminated and objects re- 
moved from a horizontal plane receive 
much less illumination. There are 
two reasons : first, the light has 
farther to travel from its source, and 
second, the intrinsic brilliancy in this 
direction is diminished very consider- 

While the various high efficiency 
lamps using metallic filaments are not 
destined soon to replace the carbon 
filament among general consumers, a 



March, 1908 

glance at their light distribution 
curves will be interesting. The same 
cut shows curves for bare tungsten 
and tantalum lamps in which the dis- 
tribution is much poorer than for the 
standard carbon filament lamps. In 
fact, they are so unsuited for use as 
a bare illuminant that it is necessary 
to employ reflectors to change the dis- 
tribution curve. Properly, the tung- 
sten lamp may not be considered as an 
illuminant per se. It requires the 
combination of lamp and reflector to 
give the proper distribution of light. 

Legal Notes 



Proof that an -electric light lamp 
suspended over a street fell and in- 
jured a traveler because the rope 
holding it broke raised a presumption 
of negligence on the part of the light 
company, and the burden is on it to 
show the contrary. Louisville Light- 
ing Co. v Oldens. Court of Appeals 
of Kentucky. 105 Southwestern 435. 


A company which has obtained 
from a city a franchise to furnish 
light for its inhabitants cannot object 
to the grant of a similar right to some 
one else or to the furnishing light by 
the city itself to its citizens. Crouch 
v. City of McKinney. Court of Civil 
Appeals of Texas. 104 Southwestern 


A lineman, engaged with others in 
removing wires from poles, who was 
injured by the falling of a pole on 
which he was at work, caused by its 
being rotten beneath the sidewalk in 
which it was planted, cannot be held 
to have assumed the risk from such 
danger under the circumstances ex- 
plained. Munroe v. Fred T. Ley & 
Co. U. S. Circuit Court of Appeals. 
156 Federal 468. 


Where a city owning an electric 
light plant has a surplus of electricity 
remaining after discharging its public 
duty, it may expend current funds to 
put that power in use so as to supply 
electricity to its citizens for private 
use. Crouch v. City of McKinney. 
Court of Civil Appeals of Texas. 104 
Southwestern 518. 



In an action by a lineman employed 
with others in removing electric light 

wires from the poles on which they 
were strung to recover for an injury 
caused by the falling of a pole on 
which he was at work, it was shown 
that the cause of the injury was the 
negligent method of doing the work; 
that the act of negligence which was 
the immediate cause of the falling of 
the pole was done by a workman by 
direction of one of two men who 
were standing on the ground, and not 
working with their hands, but giving 
directions to the workmen. Held, 
that such evidence was sufficient to 
entitle plaintiff to go to the jury on 
the question whether or not such men 
were, or either of them was, "en- 
trusted with and exercising superin- 
tendence and whose sole or principal 
duty was that of superintendence," so 
as to render the defendant, as em- 
ployer, liable for his negligence un- 
der the Massachusetts employer's 
liability act (Rev. Laws Mass. c. 106, 
§ yi). Munroe v. Fred T. Ley & Co. 
L\ S. Circuit Court of Appeals. 156 
Federal 468. 


Where plaintiff, an electric lineman, 
was severely burned through defend- 
ant's negligence by a heavily charged 
electric wire, and suffered and would 
continue to suffer indescribable pain 
in consequence of the injury, a verdict 
of $30,000, sustained by the trial 
court, will not be set aside on appeal 
as excessive. Reeve v. Colusa Gas ^ 
Electric Co. Supreme Court of Cali- 
fornia. 92 Pacific 89. 


An experienced electric lineman 
was not negligent, as a matter of law, 
in working near wires which he did 
not know were charged, without in- 
quiry as to the current of electricity ; 
it being the duty of the electric com- 
pany to use reasonable care to so con- 
trol and manage the operation of the 
system and the place where plaintiff 
was put to work that the wires should 
be free from dangerous currents un- 
der the company's control while the 
work was in progress, or to give 
plaintiff necessary warning and in- 
structions to enable him to avoid the 
danger, in so far as it was reasonably 
possible and compatible with the na- 
ture of the work. Reeve v. Colusa 
Gas & Electric Co. Supreme Court 
of California. 92 Pacific 89. 


Plaintiff, an experienced electric 
lineman, was given a general order to 
assist in transferring a transtormer 
from one pole to another. He had 
been informed that there was neces- 
sity for haste, and after he had at- 
tached the crossarm on the pole, 
read\- to receive The transformer, he 
saw his fellow servants on the ground 
getting ready to throw to him the 
rope necessary to hoist the block and 
tackle to be used in raising the trans- 
former to its place, and also saw the 
foreman take the cord containing the 
transformer and run it up under the 
pole. Held, that plaintiff was not 
negligent in exceeding his instruc- 
tions in proceeding at once to the top 
of the pole to attach the pulley to hoist 
the transformer without special in- 
structions, which he was attempting 
to do when he was burned by a heav- 
ily charged wire ; plaintiff being ig- 
norant of the fact that his foreman 
intended to delay the raising of the 
transformer until after the lunch hour. 
Reeve v. Colusa Gas & Electric Co. 
Supreme Court of California. 92 Pa- 
cific 89. 


In an action for an injury to a 
clerk in a store, caused by the fall of 
a lamp maintained by defendant elec- 
tric light company, and hung by a 
hook screwed through a thin board, 
it was improper to exclude a question, 
asked a carpenter testifying for plain- 
tiff, as to where hooks are usually 
placed when attached to ceilings; the 
obvious intention of the question be- 
ing to show that the hook should have 
been screwed into a joist. Fish v. 
Waverly Electric Light & Poicer Co. 
Court of Appeals of Xew York. 82 
Xortheastern 150. 




One seeking to restrain a city own- 
ing and operating an electric light 
plant to light its streets from selling 
electricity to private persons for light- 
ing must show that the city did not 
sufficiently light its streets, and that 
it was financially able to extend its 
system for lighting its streets, since 
the city, after discharging to the best 
of its ability its duty of lighting the 
streets, could sell its surplus power 
to private citizens for lighting. 
Crouch v. City of McKinney. Court 
of Civil Appeals of Texas. 104 
Southwestern 518. 

New Type of High-Speed Steam Engine 

The American-Ball Angle Compound 

ANEW and interesting type of 
high-speed engine is described 
in the following article for the 
first time in this paper. 

This engine marks a distinct epoch 
in the development of practical steam 
engines, and is the culminating work 
of a life-time spent in developing and 
perfecting high-speed engines. 

This latest engine is the joint pro- 
duction of Mr. F. H. Ball, the well- 
known engine designer, and his son, 
Mr. F. C. Ball. 

In the early days of high-speed en- 
gines there was great similarity in the 
valve gears of all makes, but in later 
years there has been a divergence in 
the line of development followed by 

engines and the greater liability of 
interrupting service more than offsets 
the small gain in efficiency that may 
be realized from a multiplication of 
the valves and valve mechanism, and 
that where high efficiency is desired, 
a much better plan is to use a com- 
pound engine of simple design be- 
cause it is vastly more economical of 
steam than any simple engine even 
with the most complicated valve gear. 

Mr. Ball and his son have con- 
sistently held to this view and have, 
sought to attain the extreme of sim- 
plicity and fewness of parts. The 
well-known duplex-compound engine 
is in this line of development, and 
now this newest engine, called the 
angle-compound, is another step in the 
same direction. 

The general plan of combining a 


designers of this class of engines. 
Some have sought a refinement of 
efficiency by the use of complicated 
valve gear. Others have claimed that 
the increased cost of maintenance of 
complicated valve gear on high-speed 

horizontal engine and a vertical en- 
gine so that both shall work on the 
same crank pin is not new. There 
are conspicuous examples of this gen- 
eral type in the giant engines installed 
in the traction power-houses of New 

York City, but engineers do not seem 
to have appreciated the many desir- 
able features of this arrangement for 
high-speed engines of the single-valve 

It is well understood that with re- 
ciprocating engines the question of 
counterbalance becomes increasingly 
important and serious as the speed is 

A very erroneous idea is somewhat 
prevalent to the effect that recipro- 
cating engines may be counterbal- 
anced so that the thrusts of inertia are 
neutralized. Nothing could be fur- 
ther from the truth, because a coun- 
terweight attached to the heel of a 
crank merely transfers the unbalanced 
thrust from the plane of the recipro- 
cating parts into a plane at right 
angles to it. Thus in a locomotive the 
counterweight in the driving wheel, 
which is absolutely necessary to keep 
the engine from "nosing" violently at 
high speeds, simply transfers the un- 
balanced thrust to a vertical plane. 
Recent experiments with a locomo- 
tive testing equipment have shown 
that at high speeds this vertical thrust 
becomes so great that when the coun- 
terweight passes over the shaft, the 
wheels, with the weight of the engine 
on them, are actually lifted clear of 
the track. This seems incredible, but 
has been abundantly demonstrated. 

The casual observer of high-speed 
engines does not understand that 
with a horizontal engine the inertia 
thrust of the reciprocating parts 
which will rock a foundation badly 
will, if transferred into a vertical 
plane, be easily resisted by the same 
foundation. The engine then has no 
rocking tendency and, therefore, 
seems to be balanced. 

The usual practice of engine de- 
signers is to counterbalance to the ex- 
tent of transferring the largest part 
of the horizontal thrust to a vertical 
plane, leaving only such an amount of 
horizontal thrust as will be safely re- 
sisted by the usual foundation. 

Keeping this in mind, it is evident 
that by combining a horizontal and 
vertical engine on the same crank pin, 
the total amount of horizontal thrust 
may be neutralized by counterbal- 
ance; and when the counterweight is 
in a vertical plane, it is opposed by 
the reciprocating parts of the vertical 
engine so that at four points of the 
stroke a perfect balance is realized, 
and between these four points there is 
no position of the crank when any 
appreciable unbalanced condition is 




March, 1908 

In the angle-compound engine here- 
with illustrated, the conditions for 
perfect balance are brought about by 
making the low-pressure piston a very 
light, conically-shaped steel structure 
of about the same weight as the ordi- 
nary cast iron piston in the high-pres- 
sure cylinder. The low-pressure en- 
gine is made the vertical engine be- 
cause it is thought desirable to have 
the larger piston rest on the piston 
rod rather than to drag in a horizontal 

It will be seen by reference to the 
several views of the engine that the 
high-pressure valve is driven by the 
usual valve gear and shaft governor, 
and the low pressure by an eccentric 
which is enclosed in an oil-tight cas- 

water drip from mingling with the oil 
of the circulating system. The water 
drip from all these stuffing boxes is 
carried off by concealed piping so 
that the engine is never untidy in ap- 

A new departure has been made in 
this engine in the arrangement of the 
crosshead and guides, which are of 
the bored type. It will be noticed 
that the crosshead is a single piece 
without the usual adjusting shoes, 
while the guides are made adjustable. 
These guides are carried in bored 
seats, and a projection from the back 
of the guide fits between the supports 
so as to resist and thrust. One of 
the guides is secured to the support 
by screws, and is only adjustable by 


ing and connected with the oil-circu- 
lating system of the engine. From 
this eccentric a rod drives direct to 
the low-pressure valve stem, which is 
guided by a cross-head carried in 
guides as shown, and this cross-head 
and guides are also included in the 
oil-circulating system of the engine. 
The oil-circulating system is similar 
to that used on the American-Ball en- 
gines, with which engineers are fa- 
miliar, except that the oil is pumped 
directly to the gravity storage tank 
on the low-pressure frame, which is 
kept constantly overflowing by the 
supply delivered to it from the pump. 
A double stuffing box on the valve 
stem and bulk heads and stuffing 
boxes on both piston rods prevent the 

means of shims, but the other has a 
pair of screws at each support to pro- 
vide for delicate adjustments, and the 
guide is securely held against these 
adjusting 'screws by a bolt that locks 
the adjustment securely when set up. 

It is, of course, understood that the 
crank pin is double the usual length, 
and that the connecting rods are 
placed side by side on this double 
length pin. This completes the gen- 
eral description of this new and in- 
teresting engine. 

The builders' claim that this type 
of engine runs more smoothly and 
with much less strain and shock than 
any other form of high-speed engine 
because of its perfect balance, and be- 
cause it has four small impulses on 

the crank at each revolution instead 
of two large ones would seem to have 
every justification. They have re- 
cently installed in their own power 
plant one of these new engines of 
160 h.p., 1 1 -in. stroke, running about 
300 rev. per min., direct connected to 
one of their generators. This engine 
has no special foundation, except the 
concrete floor of the building, and has 
not a single foundation bolt. The 
writer stood a new full length pencil 
on its end on the horizontal cylinder 
head and then on the vertical cylinder, 
and there was not vibration enough to 
disturb the delicate balance of the 
pencil even with a fluctuating load on 
the engine. It is apparent, also, that 
this is the only form of reciprocating 
engine that can really be counterbal- 
anced, and it is therefore better 
suited to high speed than any other 
type of reciprocating engine. 

It is evident, also, that the floor 
space of this engine is very small. 
Since a horizontal engine carries a 
vertical engine of the same power, the 
power for a given floor space is, there- 
fore, doubled. This also reduced the 
cost of foundation to the same extent, 
so that practically half the foundation 
is saved ; besides, the perfect balanc- 
ing of the engine makes the founda- 
tion problem such a very simple one, 
that where concrete floors are used no 
further foundation is ordinarily re- 
quired, thus saving the entire cost of J 

For large powers these engines, 
combined in pairs with the generator 
or belt-wheel between them, make an 
exceedingly compact unit. In these 
combinations the engines are used as 
double-compounds when run non- 
condensing, or where condensing 
water is available, as four-cylinder 
triple-expansion engines. In the lat- 
ter case one horizontal engine is the 
high pressure, the other horizontal 
the intermediate pressure and the two. 
vertical engines combined are the low 
pressure, thus giving a large area of 
low-pressure piston without using any 
pistons of very large diameter. 

Since the normal speed of an en- 
gine of this kind is high, the cost of 
the generator and the amount of floor 
space are both greatly reduced. 

Large Electrical Machines Bviilt 
at West Allis WorKs, Allis- 
CHalmers Company- 
Several months ago steps were- 
taken to stock and equip two of the 
great machine shop units at the West 
Allis works of Allis-Chalmers Com- 
pany, known as shops 5 and 6, for 
the building of large electrical ma- 
chines, particularly those intended for 
direct connection to the various forms 
of prime movers which constitute a 

March, 1908 



large part of the product of this great 

Heretofore the large electric gen- 
erators for driving by gas engine, 
steam turbine, water-wheel or Corliss 
engine have been built exclusively at 
the company's works in Cincinnati, 
Ohio. The construction of a part of 
these large machines in Milwaukee 
gives much needed room for the man- 
ufacture of motors, transformers and 
small generators at Cincinnati and 
does away with the former necessity 
of shipping heavy engine shafts to 
the electrical works to be fitted and 
keyed to the rotors of engine type 
generators. On the other hand, the 
West Allis works are better equipped, 
through large experience in the build- 
ing of big equipments, to handle the 
heavy parts. 

To indicate the type of work which 
has already been turned out at West 
Allis, shipment was recently made of 
two 2000-kw., 6600- volt, 25-cycle, 3- 
phase alternators which were sent to 
the Homestead works of the Car- 
negie Steel Company. These alter- 
nators are for direct connection to 42 
by 54 in.' Allis-Chalmers twin tandem 
gas engines, also products of the 
West Allis works. 

A third alternator was recently 
shipped to the central furnaces of the 
American Steel and Wire Company. 
This machine was a 1000-kw., 13,200- 
volt, 25-cycle, 3-phase unit, designed 
for direct connection to a 34-in. by 42- 
in. Allis-Chalmers gas engine. These 
three generators were the first to be 
completed and shipped from the new 
shops, and others, including a 6500- 
kw. unit, will follow in rapid succes- 
sion during the next few weeks. 

overhead transmission of over 2500 

In the central distributing station 
the voltage will be stepped down to 
3400 volts by nine water-cooled trans- 
formers of 1350-kw. capacity each, at 
which potential it is to be distributed 
throughout the city by both overhead 
and underground cables. 

It is interesting to note that from 
the nearest railway station the entire 
power apparatus for the Yawozo sta- 
tion will have to be transported on 
specially constructed wagons driven 
by oxen. 

The Naiko River is normally 40 ft. 
in depth, but in the rainy season the 
river often rises to 40 and 70 ft. 
above low water mark. This rising 
characteristic of the river will neces- 
sitate the building of a specially de- 
signed dam to take care of the high 

Electrical Equipment of Hydro- 
Electric Plant 

The General Electric Company is 
furnishing complete electrical equip- 
ment for a hydro-electric plant in 
Xagoya, Japan, a city with a popula- 
tion of about 250,000. situated some 
300 miles from Yokohoma. The 
main generating station will be built 
at Yawozo, on the Xaiko River, where 
power will be generated at 6600 volts 
by four three-phase, 2500-kw., 60- 
cycle, 360 rev. per min. water-wheel- 
driven generators. The generator 
voltages will be stepped up to the line 
voltage of 60,000 volts by 12 water- 
cooled transformers of 1000-kw. ca- 
pacity each, and transmitted 30 miles 
to the main substation just outside of 
the city of Xagoya. Here the line 
voltage is to be stepped down to 
11,000 volts by nine water-cooled 
transformers of 1350-kw. capacity 
each, and transmitted underground 
to the distributing station through 
triple-conductor, lead-armored cables, 
the city ordinances prohibiting an 

New Madeline Furnace of the 

Inland Steel Co., Indiana 

Harbor, Ind. 

Since blowing in the new Madeline 
blast furnace at Indiana Harbor last 
August, the Inland Steel Company 
has produced, its own pig-iron for use 
in steel making, instead of buying 
the iron from outside, as was for- 
merly the custom. The new furnace 
has a nominal capacity of 400 tons of 
pig-iron per day, and the equipment 
of the plant auxiliary to the furnace, 
as well as the facilities for handling 
the blowing engines, power equip- 
ment, etc., are all of the most ap- 
proved standard design and latest 

The new plant is located .alongside 
of the company's steel plant, and a 
portion of the power generated in the 
newly installed power station is 
transmitted to it for use in operating 
various machinery. The works have 
a protected harbor 300 feet wide for 
receiving raw material, a good dock 
1000 feet long and plenty of Lake 
Michigan water for cooling purposes. 

Ore is taken directly from lake 
boats by means of electrically-driven 
hoists and carriers and deposited in 
storage piles or bins. Scale cars on 
motor-driven trucks handle ore be- 
tween the bins and the furnace. The 
furnace proper is 85 ft. high, with 
a diameter below the hearth top of 
20^2 ft. As far as possible the fur- 
nace filling and distribution are done 

The open-hearth plant is located 
across the tracks of the Lake Shore 
& Michigan Southern Railway, and a 
concrete-lined hot metal tunnel, built 
under the railroad tracks, communi- 
cates with it. 

Coke is stored in a 60-ft. bin, and 
six bins 91 ft. 9 in. long hold lime- 
stone and ore. The cast house is 108 

ft. long, with an additional length of 
104 ft. for claw mixing and ladle- 
drying purposes. 

The boiler plant, situated directly 
back of the furnace, comprises eight 
500-h.p. Sterling boilers, each with 
individual stacks. There are two 
pairs of Allis-Chalmers Vertical Long 
Crosshead Furnace Blowing Engines 
installed, whose steam cylinders are 
44 in. and 84 in. in diameter and air 
cylinders 84 in. and 84 in. in diameter, 
with 60-in. stroke. These engines are 
designed to operate either singly or in 
compound condensing pairs. The 
steam may pass through a reducing 
valve when the low pressure side is 
run singly. 

The power plant is further equipped 
with three Allis-Chalmers cross com- 
pound Corliss engines, each direct- 
connected to a 550-kw. direct-current 
generator of the same build. The 
engines are uniform in type, having 
cylinders 20 in. and 42 in. by 42 in. 
stroke. Power is transmitted from 
the power-house to a distributing sta- 
tion at the steel mills, a distance of 
about 1200 ft., by conducting cables 
supported on steel towers, to the 
property line of the blast furnace 
plant and from this point by under- 
ground conduit beneath the interven- 
ing tracks. 

In addition to the engines already 
enumerated there is installed in this 
plant an Allis-Chalmers horizontal 
cross compound Reynolds-Corliss 
pumping engine used for hydraulic 
transmission in the mills, operating 
lifts, stands and other handling ap- 
paratus. The water end of this unit 
has a capacity of 800 gal. of water 
per minute against a working pres- 
sure of 500 lbs. per square inch under 
severe continuous service. Mr. 
Arthur G. McKee, engineer, of Cleve- 
land, designed and superintended' the 
erection of the plant. 

"Western Portable Instruments 

A low-priced line of portable alter- 
nating-current instruments has lately 
been placed on the market by the 
Weston Electrical Instrument Com- 
pany, Newark, N. J. The voltmeters 
range from 75 to 750 volts, direct 
reading, ammeters from I to 300 am- 
peres, direct reading, and the milli- 
ammetcrs range from 75 to 750 mil- 

These instruments have no dis- 
cernible working error, practically no 
temperature correction, are independ- 
ent of the frequency of the circuit and 
possess closely uniform scales. One 
of the strongest points of the high- 
grade Weston instruments is the 
dead-beat character of their indica- 
tions, and in this respect the low- 
priced instruments are fully as good. 



March, 1908 

RernorKable Performance of an 
Induction Motor 

The general sturdiness and endur- 
ance of the induction motor under ad- 
verse conditions is proverbial, and the 
following incident serves to illustrate 
the reason for their universal popu- 

A five-horse power, 440-volt, squir- 
rel-cage General Electric induction 
motor was belt-connected to a cen- 
trifugal pump in the quarry of the 
G. H. Perry Co., Sioux Falls, S. Dak., 
the capacity of the pump being 158 
gal. per min. — lifting same about 45 
ft. The motor was operated continu- 
ously during the rainy season, and 
was often allowed to run without at- 
tention during the night. 

One Sunday morning an operative 
noticed that the quarry pit was full of 
water, the motor being partly sub- 
merged. The necessity of clearing 
the pit of water being evident, the 
current was turned on. To the sur- 
prise of all, the motor came up to 
speed and carried the load until the 
pit was clear, apparently none the 
worse for its prolonged bath. On 
examining the motor the next day, 
that portion of the paper pulley which 
had been under water was found to 
be softened and considerably warped 
out of shape. This motor gave excel- 
lent service until two months later, at 
which time it was completely de- 
stroyed in a fire which consumed sev- 
eral of the company's buildings. 

All TogetHer for a Bigger, Brig'Hter 
and Busier Marion 

The Marion Light & Heating Co., 
Marion, Ind., has recently put into 
successful operation one of the best 
new business schemes evolved in 
many years. Here it is : 

The Marion News Tribune, the im- 
portant daily of that thriving city, ap- 
peared one morning as an electrical 
edition with a special electrical sup- 
plement of eight pages. The front 
page showed a large night view of 
Washington Street, the main business 
street of the city, and gave an ac- 
count of a banquet of the employees 
of the lighting company under a spe- 
cial head. 

The leading business houses were 
featured in electric signs — free ad- 
vertising, and the industrial plants 
utilizing electricity exclusively were 
also played up splendidly in half- 
tones and in interviews of their pro- 
prietors setting forth the advantages 
of the electric drive. 

The local paper has probably found 
this special edition profitable, as it 
contains a large number of extra ad- 
vertisements of the electrical and al- 
lied interests. 

The slogan of the company — "a big- 
ger, brighter and busier Marion" — 

has aroused the citizens of that city 
to largely co-operate with the earnest 
efforts of the lighting company to 
give Marion a greater commercial 

We are of the opinion that this 
clever stroke of business boosting 
hardly cost the local lighting company 
more than the distribution of a 
monthly flyer. It did, however, cost 
some thought on the part of S. H. 
Smith, the energetic superintendent 
of the company, and E. T. Hollings- 
worth, manager of its new business 
department. They are to be congrat- 

General Electric Earns $9, 800,000 

The report of the General Electric 
Co. for the year ended January 31, 
1908, will be made public some time 
in May. The statement will show a 
gross business of approximately $70,- 
000,000, which compares with $60,- 
071,883 in the preceding fiscal year. 

Had it not been for the depression 
in the latter part of 1907, the com- 
pany would have made an even better 

The ratio of profits to business 
billed was about 14 per cent, or about 
the same as in the preceding year, 
which will bring the net earnings on 
the $70,000,000 of business billed up 
to approximately $9,800,000, an in- 
crease of about $1,400,000 as com- 
pared with the fiscal year ended Jan. 

31. T 907- 

The following table gives the 

amount of business billed, profits ap- 
plicable to dividends and the percent- 
age of profits to gross of the General 
Electric Co. over a series of years, the 
figures for the fiscal year ended Janu- 
ary 31, 1908, being estimated : 
Year ended January 31st: 






*$9, 800,000 






competition and higher operating 
costs in general over the last few 

It is well known that the business 
of the General Electric Co. this year 
will not be as large as in the preceding 
fiscal year and, naturally, earnings 
are expected to show a falling off. 
However, the General Electric Co. 
will have the advantage of lower 
priced copper, etc., in the current year, 
and a reduction in operating costs 
would not be at all surprising. 

The current depression in business 
did not begin materially to affect the 
shipments of the General Electric Co. 
until January of this year. In that 
month it is understood that shipments 
aggregated about $4,400,000, or at 
the rate of $53,000,000 a year, as com- 
pared with $70,000,00 for the full 
fiscal year ended January 31, 1908. 

President Coffin in his last annual 
report called attention to the fact that 
the sales billed for the first two 
months of the fiscal year ended Janu- 
ary 31, 1908, were more than 50% 
greater than in the corresponding pe- 
riod of the preceding year. Also that 
the total stock issued and subscribed 
aggregated $65,134,300, and that there 
had been authorized but not issued or 
subscribed $14,819,866, making a to- 
tal outstanding and authorized capital 
of $80,000,000. 

Shortly after the annual report was 
made public, the General Electric 
directors voted to offer stockholders 
approximately $13,000,000 five per 
cent, convertible bonds at par. These 
bonds have all been sold. In view of 
the heavy falling off in business, no 
General Electric financing is to be ex- 
pected this year. 

The General Electric Co. is now 
operating about 50% of its normal ca- 
pacity, and its consumption of copper 
is about one-third of normal. 

The company has been buying cop- 
per, but its purchases for some time 
past have been of a hand-to-mouth 
character. — Wall Street Journal. 

* Approximated. 

The above table shows a remarkable 
expansion in the business of the Gen- 
eral Electric Co., but the decrease in 
the ratio of profits to gross cannot be 
regarded as a favorable development. 

It would appear that while the 
gross business of the company has 
shown an extraordinary increase, 
there has accompanied it an increase 
in operating costs which have oper- 
ated against a proportionate expan- 
sion in net profits. 

One explanation advanced for this 
showing is the high prices for copper 
and other material entering into the 
manufacture of electrical equipment, 


One of our esteemed contempo- 
raries furnishes the following in- 
formation. The extract is given in 
full : 

Bad joints may be considered as 
the unpardonable sin among electrical 
men, especially on circuits carrying 
small volume currents. Even with 
care, however, they will sometimes 
develop, as there are so many causes 
that may produce them. Every bind- 
ing screw, every fuse or cut-out, 
every connection of whatever nature 
may be considered as a source of bad 
joints trouble and treated accordingly. 
Xeedless to say, every line joint 
should be soldered or sleeved. — Elec- 
trical Record. 


Volume XXXIX. Number 4. 
$ 1 .00 a year ; 1 5 cents a copy 

New York, April, 1 908 

The Electrical Age Co. 
New York and London 

Published monthly by The Electrical Age Co., 45 E. 42d Street, New York. 

J. H. SMITH. Pres. C. A. HOPE, Sec. and Trcas. 


Telephone No. 6498 38th. Private branch exchange connecting all departments. 
Cable Address — Revolvable. New York. 


United States and Mexico, $1.00; Canada, $1.50. To Other Countries, $2.50 


Insertion of new advertisements or changes of copy cannot be guaranteed for the following 
issue if received later than the 15th of each month. 


Business and Editorial Departments, 33 Bedford Street 

Strand, W. C. 

London Telephone, Post Office. 3807 Central. Private branch exchange. London Cable 

Address — Revolvable, London. 
All remittances to London Office should be crossed London and County Banking Company. 


Subscriptions may be sent to the following branch offices or agencies, where they will re- 
ceive the same careful attention as at the home office in New York : 

124, Queen St., Melbourne 33, Loop St.. Cape Town 

359, George St., Sydney Unter den Linden, 5, Berlin 

61, Main St., Yokohama Nevsky Prospect, St. Petersburg 

23, Esplanada Rd., Bombay 31, bis rue de Fauburg Monlmartre. Paris 

General Agents for United States and Canada : The American News Company 

Electricity From Coal 

There is periodically brought before 
the public, puffed and put on the mar- 
ket, a new form of electric battery 
which, it is claimed, is about to ef- 
fect a revolution in our methods of 
power generation. Most of these are 
so obviously worthless that the in- 
vesting public pays little or no atten- 
tion to them, but occasionally a battery 
of sufficient novelty to attract the at- 
tention of technical men is devised 
and a great future predicted for it. A 
battery which consumes coal or carbon 
is sure to attract the attention of the 
public, and the claim that it will halve, 
or even reduce by 90 per cent., the 
cost of power is listened to with more 
or less credulity. It will not be amiss, 
then, to consider what claim these car- 
bon-consuming batteries have on the 
attention of the public. 

When a pound of coal is carefully 
burned it gives out sufficient heat to 
raise the temperature of about 8500 
lb. of water 1 degree cent. This, as 
demonstrated by Joule, Rowland and 
others, is equivalent to raising a 
weight of 1,000,000 lb. through a 
height of 12 ft., or 12,000 lb. through 
a height of 1000 ft. Therefore, if all 
the energy of coal were available, a 
supply of one pound of coal per hour 
would give six horse-power. Now. a 
very excellent steam engine and boiler 
give only about one-half horse-power 
for every pound of coal consumed 
per hour. Hense we get but one-half 
horse-power where we should get six. 
This extraordinary wastefulness is in- 
herent to all heat engines. A good 
modern steam engine utilizes nearly 
90 per cent, of the energy available 
from the steam that enters it, about 
10 per cent, being lost in friction and 
other incidental losses. 

In an average boiler, from every 
pound of coal consumed per hour, 
about 1.5 h-p. goes up the chimney in 
hot gases; about 0.15 h-p, remains in 
the hot and partly consumed ashes ; 

about 0.3 h-p. is radiated from the 
boiler, and about 0.3 h-p. goes off in 
unburned gases, leaving 3.75 h-p. to 
evaporate the water. Of this 3.75 
h-p., a part is wasted by radiation, 
condensation and friction in the steam 
pipes and engine. This amounts to 
about o.l h-p. Hence, if all the en- 
ergy of the steam were available, the 
engine should give about 3.65 h-p., but 
as a matter of fact it will only give 
one-half of one. In short, the engine 
will only convert into mechanical 
work about one-seventh of the energy 
of the steam, or about one-twelfth of 
the energy in the coal. 

The action of all heat engines de- 
pends on the flow of heat from one 
bod}- to another, and it is not possible 
to extract all the heat energy from a 
body without removing all of its heat. 
All the heat cannot be removed from 
a substance unless its temperature is 
brought down to 273 ° cent, below 
zero. The efficiency of a heat engine 
is directly proportional to the differ- 
ence between the temperatures of the 
incoming and outgoing fluid. Thus 
we see that good efficiency cannot be 
expected from any steam engine un- 
less the temperature of the exhaust is 
very low and that of the incoming 
steam very high. That is the reason 
why superheated steam is attracting so 
much attention. Even with a perfect 
engine exhausting at the absolute 
zero of temperature ( — 273 cent.), 
only about one-half of the energy of 
the coal would be available, owing to 
boiler losses. It thus appears that the 
boiler is so inefficient and the engine 
so defective in principle as a machine 
for producing mechanical energy that 
there is great need for a more efficient 
way of utilizing coal. 

As a contrast to the wasteful en- 
gine, consider the electric battery. 
Here conditions are reversed. Instead 
of only one-twelfth of the energy of 
the fuel appearing as useful work, 
eleven-twelfths are used. The effi- 

ciency of an electric battery may be 
90 to 95 per cent. Unfortunately the 
battery of to-day cannot compete with 
the dynamos and steam engine on ac- 
count of the great cost of the fuel it 
consumes. This fuel is almost in- 
variably metallic zinc and its cost is 
quite prohibitive. In the electric bat- 
tery the stored chemical energy of the 
fuel is converted directly into electric 
energy with only a few secondary and 
unimportant losses. With a battery of 
low resistance these losses are very 
slight. The ideal way of obtaining 
power is to cause some fuel of the 
value of coal to combine rapidly in 
an electric cell with some cheap oxi- 
dizer like air or water. How to do 
this is a problem that scientists and 
inventors have been occupied on for 
over half a century. The limited de- 
gree of sjiccess attained is evident 
from the fact that as a motive power 
the steam engine still holds undisputed 

The electric battery is primarily a 
chemical apparatus, so that to under- 
stand the principle of the battery it is 
necessary to grasp the chemical prin- 
ciples underlying its action. When 
any chemical action takes place there 
is either an evolution or an absorp- 
tion of kinetic energy. This energy 
usually appears or disappears in the 
form of heat. The combination of 
substances having a mutual affinity 
gives rise to kinetic energy, while the 
decomposition of a compound requires 
the expenditure of energy. These two 
cases have good mechanical analogues. 
Two substances having a mutual 
chemical affinity may be likened to 
the earth and some water at an eleva- 
tion. These have a mutual attraction. 
The water tends to fall to the earth's 
surface, and in falling gives up the 
energy it possessed in virtue of its 
position. This energy ordinarily goes 
to waste in heating the water, as in the 
falls and whirlpools of Niagara, but 
it may be usefully employed in turning 




April, 1908 

a waterwheel and so supply power to 
factories. In the same way chemical 
action usually goes to waste as heat, 
but there is an apparatus analogous 
to the waterwheel which enables the 
liberated energy to be collected and 
used for practical purposes. This ap- 
paratus is the electric battery, or volt- 
aic cell. An electric battery is an ap- 
paratus which enables the energy 
evolved by the combination of sub- 
stances having chemical affinity to be 
collected as electrical energy instead of 
as heat. Analogous to the decomposi- 
tion of a chemical compound is the 
separation or raising of water from 
the earth. This requires an expendi- 
ture of energy through a pump. 
Analogous to the pump is an electro- 
plating bath, in which the metal is, at 
the expense of outside energy, forcibly 
separated from its compound and de- 
posited on a foreign body. If carbon 
be caused to combine with oxygen 
there is an evolution of heat, while a 
considerable amount of energy is re- 
quired to decompose carbon dioxide 
into carbon and oxygen. The com- 
bustion of one pound of coal will raise 
the temperature of 8500 lb. of water 
I degree cent., but the combustion of 
one pound of pure carbon will only 
raise the temperature of 8080 lb. of 
water one degree. A carbon-con- 
suming battery would work more eco- 
nomically if it consumed carbon in 
the form of coke than if it consumed 
coal, because of the value of the gas 
and coal tar obtainable by distillation. 
The combination of one pound of 
carbon with 2 2 / 3 lb. of oxygen to form 
3 2 / 3 lb. of carbon dioxide is attended 
with the evolution of 8080 calories. If 
this heat is used to work a steam en- 
gine most of it will go to waste. If 
the evolved energy instead of appear- 
ing in the form of heat could be made 
to appear as electrical energy, we 
should have an ideal generator. Now, 
there is no way known in which car- 
bon can be made to combine directly 
with oxygen so as to give rise to elec- 
trical oxygen. All voltaic cells that 
we know of require a liquid to attack 
the zinc, carbon, or whatever fuel may 
be used. Thus in order to make car- 
bon and oxygen combine we must 
make the carbon take oxygen from a 
liquid compound containing that ele- 
ment. In order to make oxygen com- 
bine with carbon it must be separated 
from its compound, but in the case of 
many oxides a greater amount of en- 
erg}- is required to decompose the ox- 
ide in order to liberate the oxvsren 
than is produced by the combination 
of this oxygen with the carbon. The 
chemical action in such a battery 
therefore would not only give rise to 
no evolution of energy, but would 
actually have to absorb energy in or- 

der to take place at all. An example 
of such a battery is one using molten 
litharge or oxide of lead, where, in 
order to consume one pound of car- 
bon, 803 calories would have to be 
supplied. The oxides which give rise 
to evolution of energy are those of 
antimony, arsenic, bismuth, bromine, 
chlorine, copper, gold, hydrogen, 
iodine, mercury, nitrogen, palladium, 
platinum, selenium, silver, sulphur, 
tellurium, thallium and some very rare 
elements. Of these metals some are 
not found in nature and cost much 
to make, others give rise to so little 
energy as to be negligible and others 
are volatilized at a temperature below 
that at which they will attack carbon. 
The only ones left as worth consider- 
ing are the oxides of copper, nitrogen 
and sulphur. 

In the case of copper oxide, Cu O, 
one pound of carbon requires 13 lb. 
of oxide to effect complete combus- 
tion, and although the heat produced 
by the union of carbon and oxygen is 
8080 calories, the decomposition of 
the copper oxide requires 6321 calo- 
ries. The difference between these 
sums is 1750 calories, and represents 
the energy liberated by the reaction. 
This is only about 22 per cent, of the 
total heat available from the combina- 
tion of carbon and oxygen. 

If copper oxide occurred commonly 
in nature this process might be use- 
ful, owing to the reduction of the 
oxide into metallic copper. Unfor- 
tunately copper oxide does not occur 
commonly in nature and is too ex- 
pensive to make this practicable. 

Sulphur dioxide is a gas at ordinary 
temperatures and pressures, and sul- 
phur trioxide exists only in combina- 
tion. An oxide of nitrogen exists in 
nature in the nitrates, as, for example. 
in saltpeter. The energy available 
from this is about 88 per cent, of that 
obtained by direct combustion in 
oxygen. Notwithstanding this high 
efficiency, the cost is prohibitive on ac- 
count of the high price of saltpeter. 
The oxides of hydrogen have no effect 
on carbon. It thus appears that a 
voltaic cell in which carbon is con- 
sumed by an oxygen compound has 
no great future before it. 

Consider, now, the combination of 
carbon with other elements. There is 
a series of compounds of carbon 
known as hydrocarbons. There is a 
great number of such compounds, but 
they cannot be produced by the direct 
action of hydrogen on carbon at or- 
dinary temperatures. Whichever of 
these many compounds be considered, 
the energy liberated by the combina- 
tion of hydrogen and carbon never 
exceeds one-fifth of that required to 

produce the hydrogen by decomposing 
water. As water is our cheapest 
source of hydrogen, a hydrocarbon 
cell seems out of the question. The 
liquid hydrocarbon, dipropylene, is 
produced by the combination of one 
pound of carbon with one-fifth pound 
of hydrogen, the energy evolved be- 
ing 814 calories. Now to obtain one- 
fifth pound of hydrogen, 1.6 lb. of 
water have to be decomposed, and 
this requires the expenditure of 5840 
calories. What is true of dipropylene 
applies in various degrees to all the 

There is no other combination 
which carbon enters into which gives 
rise to the evolution of one-fifth of the 
energy evolved when it combines with 
oxygen. Combined with sulphur, car- 
bon forms carbon disulphide, and 
with nitrogen it forms cyanogen, the 
most deadly poison known. Very 
little energy is evolved by either of 
these reactions, or by the combination 
of carbon with the metals. 

It thus seems that there is no known 
way of acting on carbon so as to lib- 
erate sufficient energy to compete 
with our present wasteful method. It 
hence follows that an economical car- 
bon-consuming cell must be the re- 
sult of an entirely new discover}-. 

There can be no doubt that the so- 
lution of the problem of devising an 
economical voltaic cell for consuming 
carbon rests as much with the organic 
chemist and physiologist as with the 
inorganic chemist. It seems less 
probable that the problem will be 
solved by trying to consume carbon di- 
rectly than by trying to consume one 
of its compounds. Needless to state, 
this reaction must not require any 
heat to produce it, nor must it give 
rise to any appreciable amount of heat, 
and either the com]xnind acted on 
must be very cheap or the compound 
produced must be valuable and mar- 
ketable in practically unlimited quanti- 
ties. Tims, if the product of combus- 
tion were acetylene or gasolene, a 
chemically inefficient reaction might 
be economical from a commercial point 
of view. The same is true if the prod- 
ucts of combustion were easily con- 
verted into substances of commercial 
value, such as illuminating gas. 

Many carbon cells have been de- 
vised which are not voltaic cells, but 
heat engines. It is quite possible that 
an electric heat engine — that is. an ap- 
paratus converting the energy of heat 
into electrical energy without the inter- 
vention of chemical action — may be 
made more economical than the pres- 
ent boiler and steam engine. This in- 
creased economy might be due to the 
absence of boiler and to the in- 

creased efficiency of the engine due to 
the higher temperature at which it 

April, 1908 



might be worked, but it must be re- 
membered that all heat engines work- 
ing within moderate ranges of tem- 
perature are intrinsically inefficient. 
With such a thcrmo pile, as it is called, 
the efficiency might be double that of 
our present heat engine, but it could 
never be very great. 

It thus seems that there are serious 
limitations to the problem of utilizing 
all the energy of carbon or coal, but if 
in spite of these difficulties a really 
cheap source of energy should be in- 
vented, there will be as great a revolu- 
tion in our industries as attended the 
invention of the steam engine. 

The Delmar SHort-Circuit 

The difficulty of distinguishing a 
short circuit from a legitimate load 
has been a troublesome matter ever 
since the advent of heavy traction. A 
short circuit taking one thousand am- 
peres would cause enormous damage, 
but on a modern traction system where 
a starting train takes several thousand 
amperes, the station circuit breakers 
have to be set so high that they do 
not open unless the overload is greatly 
in excess of the latter amount. 

Hence it is not uncommon to hear 
of persistent short circuits entirely un- 
known to the station attendants, but 
involving thousands of dollars dam- 
age. This condition of affairs need 
no longer exist, thanks to the in- 
genious device described by Mr. Del 
Mar in this issue. The device is un- 
patented, and being endorsed by the 
New York Central Railroad as emi- 
nently satisfactory, should obtain ex- 
tensive use wherever large feeders are 

The application of the system to 
third rails, although not yet tried, ap- 
pears excellent. In the event of a 
train being derailed a long way from 
a substation, the short circuit cur- 
rent might not be sufficient to open 
the circuit breakers, but might set fire 
to the wreckage and cause serious loss 
of life. The possibility of this could 
be entirely obviated by the use of a 
short circuit indicating wire run along 
the third rail. If this system had been 
in use in the Paris Metropolitan sys- 
tem the terrible accident of August, 
1903, could not have happened. 

New Patent Laws in Britain 

Hitherto the patent laws of the 
United Kingdom have been quite sim- 
ilar to our own. requiring merely a 
nominal fee, no taxes, and placing no 
obligation on the patentee to work his 
patent within the term of its life. 
With the passage of the Patent and 
Designs Act, 1907, effective January 
1, 1908, the situation of the American 

inventor taking out a British patent 
is greatly altered. He is now required 
to work his patent within four years 
and to make payments during its life, 
aggregating £155 ($754.50). On the 
application for the provisional pro- 
tection, the patentee is feecd £1 with a 
fee of £3 on filing the complete speci- 
fication and a further fee of £1 at the 
granting of the patent. At the end of 
four years from the date of patent, he 
may pay £50, and at the end of eight 
years further, £100; or beginning at 
the fourth year, he may make annual 
payments of £10 each, continuing un- 
til the eighth year, when the payment 
becomes' £15, and £20 after the 10th 
year. A patent shall cease if the pat- 
entee fails to pay the prescribed fees 
within the prescribed times. 

At any time not less than four 
years after the date of a patent, any 
person may apply to the Controller 
for revocation of the patent on the 
ground that the patented article is 
manufactured and carried on exclu- 
sively or mainly outside of the United 
Kingdom, and if this be established, 
on inquiry, the patent will be revoked. 
There are sundry limitations and 
privileges arising after various con- 
tingencies. The act is retroactive 
and will apply to all patents taken out 
in 1904. 

The far-reaching effect of the new 
patent laws can hardly be measured 
at the present time, but it can safely 
be said that ultimately it will compel a 
large number of German and Ameri- 
can manufacturers to build plants in 
England to hold their trade, as other- 
wise their patents will be revoked and 
British manufacturers will seize them. 
It will be interesting to see how the 
clause providing for the revocation of 
a patent upon an article manufactured 
mainly outside of the United King- 
dom will work out when the courts 
come to take the matter in hand. 

alcoves, so that from six to twelve 
lights could readily be installed on each 
table. None of the fixtures has been 
installed, as it is understood that the 
Building Committee is undecided as 
to how the library would be finally 

Two circuits have been run to each 
alcove for special outlets for lighting 
around the book-cases, but they have 
not yet been put in use. Unquestion- 
ably this part of the illumination 
should be remedied at once and a com- 
plete outfit of lamps should be installed 
on the reading-tables. 

The LigKting of the Institute 

In the last issue of The Electric 
Age we had something to say about 
the bad lighting of the library in the 
United Engineering Building. What 
was said of the lighting is indubitably 
true. It appears, however, that the 
lighting plans of the consulting en- 
gineer were not properly followed out, 
and the faulty lighting cannot there- 
fore be imputed to them properly. 
The room is amply wired for the plac- 
ing of sufficient lights and only some 
tardiness or indecision on the part of 
the proper committee has kept the 
library in its miserably lighted condi- 
tion during the past year. 

At the time of building a great num- 
ber of floor and ceiling outlets were 
installed, both in the center of the li- 
brary floor and in the center of the 

A. Half Decade of Steam Turbine 

We publish at another page the first 
authentic figures on steam turbine 
sales covering a period of about five 
years elapsing since the commercial 
exploitation of the steam turbine in 
this country. The aggregate kilowatts 
of Curtis 'turbines is 1,073,695, as 
against 640,700 kw. of Westinghouse 
turbines. Definite figures are not fur- 
nished us by the Allis-Chalmers Co., 
the only other important manufactur- 
ing company, but we should estimate 
their sales at under 200,000 kw. They 
have come into the field later than the 
other two. 

A greater capacity of steam turbines 
of the vertical type has been sold than 
of both types of horizontal turbine, 
though it is not to be inferred from 
this statement that engineers generally 
show preference for the vertical type. 

The collaborator of the Westing- 
house figures has seen fit to omit the 
exact number of 'machines in each 
classification, but since this figure was 
necessary to get the average kilowatt 
which is given in the table, it is just 
as certainly easy to approximate the 
number by the inverse calculation. 
We make it 525, approximately. The 
number of vertical machines sold is 
1096, or about twice as many. The 
Allis-Chalmers Co. has probably sold 
about 200 turbines. All counted, then 
there are about 1800 machines sold, 
and about 1500 in service. 

The General Electric Co. has sold 
243 turbines for industrial and mis- 
cellaneous purposes and the Westing- 
house Company has sold about 235, as 
closely as we can reckon from the fig- 
ures they give out. It would appear, 
therefore, that the turbine business has 
been about evenly apportioned in the 
industrial field. " This -hows prett\ 
clearly that engineers have been about 
evenly divided in their opinion of the 
merit of the respective types. 

The total kilowatts sold is about 
t ,800,000 kw., which represents, at 
current prices, an investment of about 
$80,000,000 in this type of prime 

Central Station Distributing System 

Secondary Distribution 


Commonwealth Edison Co., Chicago 

IN the early stages of the introduc- 
tion of alternating-current systems, 
the use of 52 to 55 volt secondary 
circuits was advocated by some en- 
gineers because of the superior life of 
incandescent lamps of these voltages. 
Such a voltage was not permissible in 
direct-current work because of the ex- 
cessive amount of copper required, 
but was quite feasible in alternating- 
current systems because of the possi- 
bility of locating transformers close 
to the consumers' premises. At this 
voltage, however, it was not possible 
to supply more consumers from one 
transformer than could be reached 
from the pole on which the trans- 
former was placed without an exces- 
sive use of copper. The result was a 
system in which a large number of 
small transformers were required, 
which consumed an excessive amount 
of energy in their cores and required 
the operation of extra generating 
capacity during the light load period 
to supply their large leakage currents. 

As soon as such a distributing sys- 
tem attained such size that these items 
became too expensive a remedy was 
sought. The higher voltage lamp 
having been improved, no volt secon- 
daries were introduced and the 55 volt 
consumers were gradually changed 
over to no volts. The use of the 
higher voltage increased the range of 
distribution so that a single no volt 
transformer was installed to replace 
several 55 volt transformers, with a 
saving in the amount of capacity re- 
quired and a very great saving in the 
core losses and leakage currents. 

Later, the availability of the Edison 
three-wire system for general secon- 
dary distribution increased the range 
of such lines by permitting the use of 
110-220 volt mains, with no volt 
lamps. With this system it is possible 
to supply consumers economically 
within a radius of 400 to 600 feet 
from the transformer. This increases 
the number of consumers which can 
be carried from a single transformer 
and thereby reduces the capacity re- 
quired per kilowatt connected. This is 
due to the fact that different con- 
sumers take their maximum demands 
at different times and the resulting 
maximum demand on the transformer 
therefore never equals the sum of the 
maximums of the individual con- 


This system of secondary distribu- 
tion is the one most generally used 
in American practice as it is the sim- 
plest and most economical for lighting 


A system of secondary mains passes 
through three general stages of de- 
velopment in expanding from a small 
to a large system. 

1. — A period in which scattered 
transformers supply isolated secon- 
dary mains not interconnected with 
other transformers. 

2. — A period in which the mains 
from adjacent transformers grow to- 
gether along principal thoroughfares 
where they may be connected to each 
other but intersecting few other sec- 
ondary mains of importance. 

3. — A final stage in which secon- 
dary mains are required on nearly all 
streets and are therefore joined into a 

The first period is that found in 
residence and other outlying territory 
not fully built up. When a new con- 
sumer is to be connected in such a 
territory the problem is — Shall a 
transformer be installed or the nearest 
secondary main extended to the prem- 
ises? The installation of a trans- 
former involves an investment and an 
operating expense, due to its core loss. 
The extension of the secondary main 
also involves an investment in con- 
ductors and perhaps an increase in 
the capacity of an existing trans- 
former. The cost of the two alter- 
native plans being ascertained the one 
selected should be that which involves 
the least annual cost for interest, de- 
preciation and operation. For in- 
stance, assume that service is required 
for a new consumer, for a load of 
one kilowatt, at a point where there 
is no secondary main available. As- 
sume that if a separate transformer 
is installed the investment will be 
about $30.00 and that if the nearest 
secondary main is extended the ex- 
penditure will amount to $40.00. How 
shall service be given ? 

If the $30.00 investment is made 
there will be fixed charges at the rate 
of 15 per cent., amounting to $4.50 
per annum. There will also be an 
operating expense, due to the core loss 
of a one kilowatt transformer of about 
30 watts 8760 hr. per year, or 263 kw- 
hr. At one cent per kilowatt-hour, 

this amounts to $2.63 and the total 
annual cost is $7.13. 

Were the $40 expended for a secon- 
dary main the fixed charges at the rate 
of 12 per cent, are $4.80 and no oper- 
ating expense is added unless it be 
required to increase the capacity of 
the transformer then installed. If it 
is necessary that one kilowatt be added 
to the capacity of the existing trans- 
former the expenditure would be in- 
creased about $8.00, adding a fixed 
charge of $1.20 and an operating ex- 
pense of about 10 watts or 87 kw-hr. 
per year, costing 87 cents. The cost 
per year under this plan is therefore 
$4.80 in case the existing transformer 
can take the added load without 
change, or $6.87 if it cannot. In this 
case it would therefore be preferable 
to extend the secondary rather than 
to install a separate transformer. 

There is little occasion in this period 
of development to connect secondary 
mains in multiple. Where the mains 
have been extended until they meet 
each other it is usually preferable not 
to interconnect them, as the blowing 
of the fuse of either transformer 
shifts the load to the other and 
overloading it blows its fuse also ; 
and transformers are so far apart that 
they cannot share each other's load 
to any appreciable extent in case of an 
overload on either of them. 

The second period of development 
is reached when consumers become so 
closely situated that it is necessary to 
provide a secondary main along the 
entire length of a thoroughfare. This 
condition is ' usually first met along 
business streets and boulevards, and 
results in a long secondary main fed 
at intervals by transformers, but inter- 
sected by few other secondary mains 
of importance. When such a main 
has been established, it is the problem 
of the engineer to determine how far 
apart transformers should be located 
and what size of conductor should be 

The density of the load varies in 
different parts of the street and there 
are large blocks of load at particular 
points which make the problem a per- 
plexing one at best. A general solu- 
tion is usually not possible, owing to 
the widely varying local conditions. 
However, a determination may be 
made which will serve to indicate the 
approximate limits within which the 

April, 1908 



most economical arrangement of 
transformers and wire will lie, and 
from which some general principles 
of value may be deduced. 

Assuming that an overhead three- 
wire 115-230 volt secondary main 


transformers is 3000 watts from the 
assumed values in the table above. 
This goes on 24 hr. a day, 365 days 
a year or 8760 hr. and therefore 
amounts to 3X8760=26,280 kw-hr. 
At one cent this will cost $262.80, or 

Size ', Si*« * .'ironJDistance;Value 
of : Wo. of JLoss: "between! of 
yire t Tran-sf . ,'Wtts}Tra,nsf . .'Tranaf. 

35/? oa -.Value: 

Transf. : of \ 

:Vire : 

IZfo on .'Value of. Total 
Wire '.Core Lose: Cost r 
•' at U : 

! 20-15 



l§P JC.W. per 1000 Beet.. 

1 36. 

! f 263. 



$ 405. t$300.; 

: $704. 


'. 12-25 

-.2400 ; 


: 2520. 

378. '. 675.I 


: 210. 

i '669 4 

; 10-30 


600 ' 

: 2400. 

360. J1098.: 


: 202. 

1 690. 


1 6-50 



! 1950. a 



( 157. 

! 828. 





I 20-30 
I 15-40 
I 12-50 
I 10-60 

~ — r 


10O K. W. per 

1000 Fe«t 


', 4800 , ( 

; 4200, ; 

; 3900, j 

! 3600. : 


720. 1 442.1 

630. c 1098, 1 

585. U335.I 

540. i2070,l 

24-6 . 

402. 1 1175. 

340. 1 1100, 

314, 1 1059, 

305i t 1091, 

150 K.W, per 1000 Beet 

3 00 


t 20-45 
I 15-60 
I 12-75 
I 10-90 

T I 


500 < 

~i r 

1 6000, | 

I 5400' | 

I 5040. l 

» 4800 r i 

"i r 

900, i 675,1 

810, (1335,1 

756, 12070,1 

720, (3150, i 

81. I 

160. I 

246. i 

410, i 

490. I 1441. 

430, i 140». 

3 92. l 1394 

345 . ( 1475. 





I 60-15 
1 45-20 
1 36-25 
i 30-30 

1 r 

16900 J 

150 X. tf. per looo Feet 
(3-phase, 4-wire) 



i 8100 I 

I "900, 'l 

I 7560, l 

1 7200, i 

i 1 l 

1215. i 590. i 71. i 

1185. 11464.) 175, • 

1134. H791.1 215, i 

1080. 12760.1 387, i 

78S. I 2074 

691. i 2051. 

631, i 1980. 

605. i 2072- 


6ooo ft. in length is uniformly loaded $263.00. The value of transformers 
at intervals of 100 ft., what will be is 2oXi35=$2700.oo, and the fixed 
the best size of wire and proper spac- 
ing of transformers? 

The wire will be calculated to give 

two per cent, regulation from the 
transformers to a point midway be- 

The value of transformers and wire 
with different spacings will be calcu- 
lated for each of three load densities. 
Interest is taken at five per cent., de- 
preciation on wire at seven per cent, 
and on transformers at 10 per cent. 
The cost of energy at one cent per 
kilowatt-hour and the cost of weather- 
proof wire at 15 cents per pound. The 
cost of transformers and their iron 
losses are assumed as follows : 

charge's are $405.00, at 15 per cent 

per year. The value of 18,000 feet, 

1980 lb., of No. 6 weather-proof at with a spacing of about 500 ft. be 

In like manner the calculation is 
made for the other spacings of trans- 
formers at this load density and for 
the other densities of 100 kw. and 150 
k\v. per 1000 ft. Calculations are also 
made for a four-wire three-phase sec- 
ondary operating at 115-200 volts, at 
a density of 150 kw. per 1000 ft. for 
purposes of comparison. 

In case of water power the value of 
core loss should not be included and 
in large steam plants its value would 
be less than one cent per kilowatt in 
many cases. 

Table 1 shows the results of the cal- 
culation for overhead lines and Table 
2 for underground cables. The cable 
calculations are based on the use of 
single conductor paper insulated 
cable, having 4 / 32 in. insulation and an 
equal thickness of lead sheathing, with 
copper at 16 cents and lead at five 
cents per pound. 

The curves in Fig. 3 show the varia- 
tion of the various elements compos- 
ing the annual cost with a load of 50 
kw. per 1000 ft. overhead. 

Fig. 4 shows the variation in annual 
cost on overhead lines at the three 
load densities assumed, together with 
the calculations for a three-phase sec- 
ondary at 150 kw. per 1000 ft. 

Fig. 5 shows the same for under- 
ground lines. 

It is at once apparent from the 
curves in Fig. 4 that the minimum 
annual cost with overhead lines occurs 


§JJae|i>l2e & : Iron ;DTTE^" 

of IWo, of : Lor;s sance 
feole iTra/ief .:Watta -Impart 

Tranef . 

15/* or\ 
Tran e ( , 


1^ on 





^ k-.i^, per lo™ f<*>* 








24 OO 





7 950. 

* 405", 





/ 924. 
: 936 

106* , 

Size K. W 

Watts, Iron Loss .... 
Cost, Dollars 

10 15 20 25 
110 150 175 20 
100 135 175 210 

30 40 
230 260 
240 280 

Size, K. W 

50 60 
300 330 
320 360 

75 90 
380 430 


Cost, Dollars 




. , 

100 r.V. T> er 1000 ?eet 


















































350 KAf. per looo Teet 


■ ■- , 



*, 200 

r ■ : 
; 7200. • 


1 19SO| 



l 002. 





; 300 

', 6O0O, 


) 2340] 


( 490, 





; 400 

', 5400. J 


j 3850( 


; 430. 





', 500 

', 5C40. 1 

7 56. 

,• 52/XJ; 


1 392. 
.' 376. 



30-30 ' 


: 600 

1 48O0. 1 


: ?ooo : 



Calculations are made for load 
densities of 50 kw., 100 kw. and 150 
kw. per 1000 ft. of line. 

For instance, in the case of a load 
of 50 kw. per 1000 ft., a spacing of 
transformers 300 ft. apart would re- 
quire 20 15-kw. transformers for the 
run of 6000 feet. It is found that 
No. 6 B. & S. will give two per cent. 

150 K. W. per looo Feet 

(3-phase, 4-udre 

i ! ! ! ! ' ! ! — 

6 ! 90-10 ! 9900 l 200 I 9000, ; 1350, ; 2280; 

4 i 60-15 i 900C ', 3C0 { 8100. ; 1215. ; 2640; 

; 4-5-20 1 7 900 I 4O0 .' 7900. • 1165. 1 4450 • 
2/0 ; 36-25 i 7200 t 500 ', 7560, / 1134, ', 515 0; 
4/o ', 30-30 ? 6900 '. 600 j 7200. \ 1060. ,' 69501 


! 865, 



(788, ' 

23 96 


J 691. , 



' 651, 



1 605". 



15 cents per lb. is $297.00 and the tween transformers, whether the load 

fixed charges at 12 per cent, are $36.00 is 50 kw. or 150 kw. per 1000 ft. With 

approximately. The total annual cost underground lines the most econoin- 

regulation in the secondary wire at is therefore $263.00, plus $405.00, plus ical spacing is about 300 ft. between 

full load. The iron loss of 20 15-kw. $36.00, or $704.00. transformers. 



April, 1908 

It is also evident from the shape of 
the curves of annual cost that the 
percentage of variation is small when 
other spacing than the most econom- 
ical is considered. For instance, -in 
the case of a load of ioo kw. per iooo 
ft. overhead, a spacing of 300 ft. in- 
creases the annual cost but 1 1 per cent, 
over the most economical spacing of 
500 ft. and an increase in the spacing 
results similarly. In view of this con- 
dition it is possible to allow some flexi- 
bility in spacings in order to take ad- 
vantage of other considerations. For 
instance, the susceptibility of trans- 
formers to lightning and similar dis- 
turbances makes it desirable to work 
toward the longer spacings and a 
lesser number of transformers. The 
curves indicate that this can be done 
if desired without seriously affecting 
the economy. Again, it is usually de- 
sirable in building extensive secondary 
mains to anticipate an increase in load, 
by erecting a larger conductor than 
is required for immediate needs. This 
may be done and the spacing of trans- 
formers gradually lessened as the load 
increases. The cost will be slightly 
excessive at first, but decreases until a 
spacing of about 500 ft. is reached 
and then increases as spacings are fur- 
ther lessened. 

Further growth must then be pro- 
vided for by the replacement of the 
overloaded portions of the main by 
conductors of a larger size. 

The entire foregoing discussion is 
based on an assumption that the load 
is evenly distributed along the line 
throughout its length. 

Unfortunately such is usually not 
the case in practice. It is usual to 
find a portion of a secondary main 
heavily loaded and other portions 
lightly loaded owing to differences in 
the character of the neighborhood 
through which it passes. At- intervals 
a large store, church or other con- 
sumers of electricity may throw heavy 
loads upon the line. 

It is therefore necessary in practice 
to locate transformers as closely as 
possible to such large consumers' 
premises and design the main between 
them to carry the scattered consumers 
whose load is approximately evenly 
distributed. An extended secondary 
main may therefore be made up of 
several sizes of wire at different points 
with transformers having various 
spacings, depending upon the load 
density in the vicinity. 

The design of the various portions 
of such a secondary main is therefore 
likely to be difficult unless the general 
theory outlined in the foregoing is 
taken into consideration, and intelli- 
gently applied. 

The network is the last step in the 
development of a system of secondary 














S /3oo 


N /2oo 












■ fG 




















00000 20000a 

FIG. 4. 


April, 1908 



mains. It consists of a number of 
mains running at angles so that they 
cross each other and may he inter- 
connected at all intersecting points 
Transformers are located at points 
of intersection where they deliver cur- 
rent in all directions with the best 
economy of copper. The transformers 
thus maintain the full feeder end pres- 
sure at all junction points where they 
are placed, as the primary mains are 
so short that there is no appreciable 
drop in them. This is an advantage 
over a direct-current network where 
each feeder usually is connected to 
the network in only one or two places 
and the mains must have greater ca- 
pacity in order to maintain an even 
pressure throughout. 

Large networks are usually under- 
ground since this form of construction 
is commonly required by municipali- 
ties in congested sections. This form 
of construction is favorable to the 
maintenance of lines and the require- 
ments of continuous service. 

The provision of manholes of suit- 
able size for the large transformer 
units required in sections where the 
load is very dense becomes a difficult 
problem. The presence of conduits, 
gas pipes, water pipes, car tracks, etc., 
utilizes so much space that it becomes 
a physical impossibility to secure clear 
space in manholes for large trans- 
formers, except by excavating to a 
depth where drainage becomes impos- 
sible. The difficulty and expense un- 
der such conditions may make it pref- 
erable to establish a transformer sub- 
station from which low tension feed- 
ers emanate as in direct-current sys- 

These substations having no moving 
apparatus may be located in a com- 
paratively small space in a basement 
so situated that the feeders may be 
short and may not require indepen- 
dent regulation. 

The design of secondary systems is 
subject to some restriction, when in- 
ductive loads, such as arc lamps and 
motors must be served along with in- 
candescent lighting. 

The regulation of most transform- 
ers is not so good with inductive load 
as with non-inductive load, owing to 
the reactive drop in their windings. 
In case power and lighting loads are 
supplied from the same transformer 
the heavy starting current required 
by the induction motors may momen- 
tarily overload the transformer with 
current at low power factor. This low 
power-factor current drops the pres- 
sure on the secondary line for a few 
moments and causes a flickering of the 
incandescent lights. If the motor load 
is a considerable part of the total load 
the pressure remains lower while the 
motors are in operation and varies as 
the motor load changes. It is there- 

U >re necessary to install separate 
transformers for power load and for 
large installations of arc lamps where 
these constitute a majority of the load, 
if the best regulation is required for 
the incandescent lighting. 

The regulation which will be se- 
cured with a given transformer may 
be calculated for any set of conditions 
which may arise, if the impedance 
drop of the transformer is known. 

The impedance drop of a trans- 
former is that pressure applied at the 
primary terminals which will cause 
full load current to flow in the secon- 
dary when its terminals are short cir- 
cuited. For instance, in a certain 10 
kw. transformer wound for 2200-110 
volts, it was found that a pressure of 
80 volts applied at the primary termi- 
nals would cause full load current to 
flow through an ammeter connected 
across the secondary terminals. The 
impedance drop was therefore 80.0 
volts or 3.6 per cent. The impedance 
drop is the resultant of the resistance 
drop and the reactance drop just as 
in the case of an electric circuit. The 
resistance of the primary and secon- 
dary coils measured by means of direct 
current was found to be such that at 
full load the resistance drop was 1.8 
per cent, or 40 volts. 

The reactance drop is therefore — 

\/ (8oy— ( 4 o ?= 

■\/6400 — 1600=69 volts=3-i% 

In Fig. 6 let OA be the impressed 
pressure on the primary at no load. 
AB is the ohmic drop in the trans- 
former windings, which in this case is 
40 volts. This is in opposition to the 
impressed electromotive force and 
must therefore be added directly to it 
in determining what pressure must be 
impressed on the transformer in order 
to deliver its rated secondary pressure 
at full load. BC represents the induct- 
ive drop of 69 volts which must be 
laid off at right angles to AB. The 
pressure necessary to secure no at 
th e secondary at fu ll load is therefore 

\J (2240) 2J r(6g) 2 =2241 volts. With 
an incandescent lamp load of 100 per 
cent, power-factor the regulation of 
this transformer is therefore 1.8 per 

With a load of 10 apparent kw. at 
70 per cent power-factor the regula- 
tion is calculated thus : 

In Fig. 6 the impressed pressure, 
2200 volts at no load is OE. This is 
opposed by a power consuming com- 
ponent in the load of 0/7=0.7X2200 
= 1540 volts, and a wattless compo- 
nent £#=0.71 X 1 10=1562 volts. The 
ohmic drop in the transformer FF= 
40 volts and the inductive drop FG= 
69 volts. The impressed pressure at 
the primary necessary to maintain no 
volts at the secondary of the trans- 
former is therefore — 

OG= V (OH+ EFy+(E H+FGy 

OG=V(i58o) a +(i6 3 i) 2 = 
2270 volts. 

The drop at 70 degrees power fac- 
tor full load is therefore 70 volts=3.2 
per cent. At 100 per cent, overload 
this would be 6.4 per cent. With a 
motor taking current at a power 
factor of 60 per cent, or less at 
starting it is evident that incandescent 
lights supplied by the same trans- 
former will flicker whenever the motor 
is started and will burn at reduced 
candle-power while the motor is run- 
ning, unless the motor load is so small 
compared with the lighting that the 
starting current is less than the full 
load current of the transformer. 

With a load consisting of arc lamps 
the power factor is approximately 70 
per cent, and the drop at full load is as 
in the foregoing case about 3.2 per 
cent. This would be considered too 
much for satisfactory incandescent 
lighting in many cases and if so it 
would be necessary to set a separate 
transformer for the arc lamps. When 
combined with an equal amount of in- 
candescent lighting, the resulting 
power factor at the transformer is in- 
creased to about 93 per cent, and the 
regulation of the transformer is not 

The table in Fig. 7 shows some of 
the characteristics of line transform- 
ers of the sizes commonly used, which 
will be of use in making calculations. 
Improvement has been made by re- 
ducing the reactance drop in the 
smaller sizes of transformers by some 
manufacturers in recent years. The 
installation of separate transformers 
for power load necessitates separate 
secondary systems for the power con- 
sumers whose premises are in the 
same vicinity. The design of power 
"secondaries" is governed by the same 
principles that control the arrange- 
ment of lighting mains, except that it 
is permissible to allow the regulation 
to be as high as 5 per cent instead 
of 2 per cent. It usually permits 
longer runs and requires no more 
copper than is necessary for carry- 
ing capacity. In manufacturing 
districts the power load usually ex- 
ceeds the lighting, and duplicate sec- 
ondary systems are frequent, though 
not close enough together to per- 
mit interconnection to any extent. 
In mercantile districts the reverse is 
the case and the heavy lighting sec- 
ondary system is capable of absorbing 
miscellaneous power without seriously 
affecting the lighting service. The 
use of separate transformers for 
power in such sections is therefore 
not necessary except for occasional 
large consumers. 

In two-phase systems having three- 
wire single-phase lighting secondaries 



April, 1908 

two additional secondary wires are 
required for two-phase power con- 
sumers, making a five-wire service 
necessary, where light and power are 
served in the same building. 

In a three-phase system two meth- 
ods of carrying mixed light and power 
load are available. The most com- 
monly used consists of star-connected 
transformers supplying a four-wire 
main operated at 115 volts from phase 
to neutral and 200 volts across phase 
wires. Lights are balanced as nearly 
as possible on the three phases. The 
smaller lighting services may be made 
three-wire, being connected to two 
phases and neutral. Four-wire serv- 
ice is required wherever both light 
and power is to be served in the same 
building. The chief objections to this 
system are the difficulty of maintain- 
ing a balance and the necessity of in- 
stalling three transformers at each 
point where the secondary main is fed. 

In the other method, which is illus- 
trated in Fig. 8, all the lighting is 
carried on one phase by means of a 
three-wire Edison secondary. Small 
power may then be served by the in- 
stallation of one additional smaller 
transformer, and a fourth secondary 
wire operating on the open delta con- 
nection. Larger power may require 
two power transformers in addition to 
the lighting transformer. This sys- 
tem is easier to keep balanced, and 
since all the lighting is on one phase, 
permits the use of less transformer 
capacity for lighting purposes and re- 












. /roo 








vO J »>--' 















L/ASq**4 / *°' '*S& 


C/A Cut. 1 T* 



FIG. 5. 

wire that in the case of a load density 
of 150 kw. per 1000 ft. on the four- 
wire three-phase system the minimum 

di xes transformer investment and 
core loss materially as compared with 
the star-connected secondary. 

It will be noted in the calculations 
made for the most economical size of 

annual cost of such a system is 
Si 998.00. as compared with $1413.00 
for an Edison three-wire system with 
the same conditions. This difference 
is due to the fact that three transform- 

ers of small capacity are required in 
the three-phase system as against one 
in the single-phase. The saving in 
wire due to the use of three-phase 
transmission is therefore much more 
than offset by the increased cost of 
transformers and greater iron losses. 

Another advantage of the system in 
which all lights are carried on one 
phase is that the effect of the start- 
ing current of motors is noticed less 
on the lighting supplied by the large 
unit on the lighting phase than it is 
where the starting current is drawn 
from three small transformers, each 
of which carries lighting load. It is. 
therefore, possible with this system 
to carry larger power loads intercon- 
nected with the lighting than in the 
star-connected system under the same 

Where a network has been devel- 
oped, this system can. of course, not 
be interconnected with other secon- 
dary mains except those which are fed 
from the same primary phase. Under 
these conditions it is necessary to sec- 
tionalize the network so as to divide 
the load between the primarv phases. 
As the size of the mains in the network 
increases this becomes undesirable, 
and the objections to the star-con- 
nected system becomes less important. 
The four secondary conductors may 
then be changed over to a star-con- 
nected system using the three heavv 

April, 1908 



conductors which were formerly used 
for lighting" as the outer wires of the 
new system and the power wire as 
the neutral. The network may then 
be interconnected throughout and in- 
creased reliability of service thus se- 

The use of combined light and 
power secondary mains becomes de- 
sirable in an underground system as 
soon as there is a sufficient number 
of power consumers to warrant a gen- 
eral system of power secondaries in 
any locality. 

The expense of extra ducts, man- 
holes, and separate cables for power 
secondaries soon becomes prohibitive 
and it is therefore found desirable to 
combine light and power secondaries 
into one system at an earlier stage of 
development than in the case of over- 
head lines. 

The selection of the proper size of 
transformer for the supply of various 
classes of consumers is a matter of 
great importance since excess capacity 
involves idle investment as well as 
unnecessary core losses. The size of 
transformer units should therefore be 
kept as low as possible consistent with 
preservation of the apparatus and 
good regulation. 

Very few electric light and power 
consumers use their entire connected 
load at any time. In lighting there 
are always some lamps which are not 
in use at times when the principal 
part of the lighting is on, and in power 
the load is frequently less than the 
rated capacity of the motor. Where 
there are a number of motors in use 
the maximum load is rarely on all 
of them at the same time. 

Where a number of consumers are 
grouped on one transformer the maxi- 
mum demands of the various consum- 
ers do not occur simultaneously and the 
resultant maximum demand must be 
ascertained by measurements. These 
measurements may be made by means 


maximum for the entire period thus 
determined, whereas readings taken 
with an ammeter give results which 
are applicable only to the time at 
which the readings are taken. Cer- 
tain ratios of maximum demand to 
connected loads may be established by 
a series of such measurements for 
the various class of consumers for 
which it is necessary to select trans- 
formers. These ratios may then be 
applied with reasonable certainty to 
the selection of transformers for new 

For instance, it has been found in 
the City of Chicago that in store light- 
ing the maximum demand for window 
lighting, signs and other display 
lighting is practically ioo per cent, of 
the connected load. The demand on 
interior store lighting is 75 to 80 per 
cent. There are usually two or three 
nights in the week in which the de- 
mand will be less than this. 

In residences where the connected 

vidual consumers is a much larger per- 
centage of the total. 

In the case of churches and similar 
public buildings capacity must be pro- 
vided for the illumination of the 
largest room in the building together 
with the necessary hallways and cor- 
ridors. This usually requires capacity 
for at least 75 per cent, of the con- 
nected load. 

In theater lighting allowance may 
be made for the use of border and foot 
lights of several colors which are not 
used simultaneously and for the fact 
that the stage and auditorium are not 
lighted simultaneously except for a 
very few minutes at a time. In a small 
theater the ratio may be from 70 to 85 
per cent, while in a large theater it 
frequently runs as low as 50 per cent. 

Where several classes of buildings 
are fed by one transformer, the ca- 
pacity must, of course, be determined 
by taking each class into consideration 
separately and thus arriving at an av- 
erage ratio for the whole. 



//S/O*. 71S 

//£ vb*-7s 


//S-220 vbi. r 
fig. 8. 


230 Vo^r Zy/pSjS- 

K. W. 






at 100 Per 

Per Cent. 

Per Cent. 

Cent., P. F. 




























- 2. 








' 1.8 




~ 1.7 




1 .65 






1 55 

3.5 r 

7 1.6 





FIG. /. 

of a split ring current transformer and 
ammeter or by the installation of a 
Wright demand indicator. The use 
of the demand indicator is preferable 
as it may be left in circuit throughout 
any desired period and the absolute 

load is 50 lights or more, the average 
maximum demand of a group of resi- 
dences is 15 to 20 per cent, of the con- 
nected load. Individual residences 
may have occasional maximums of 30 
to 50 per cent, for which some allow- 
ance should be made in selecting 
transformer capacity. The size of the 
transformer should be such that it will 
carry the occasional high maximum 
of the largest individual consumer to- 
gether with the average maximum of 
the other consumers on the trans- 
former. Oil transformers may safely 
be permitted to carry 25 to 50 per 
cent, overload occasionally in such 

Small residences and apartments 
having connected loads of 40 lights 
or under average about 20 per cent, 
of the connected load, with 25 to 30 
per cent, as an occasional maximum. 

In general, a higher ratio must be 
used where there are but two or three 
consumers on a transformer than 
where there are 10 or more consumers. 
as the occasional maximum of indi- 

The selection of transformers for 
power consumers is a more difficult 
task, as the maximum load may vary 
greatly from day to day or from 
month to month. Consumers having 
but one or two motors generally re- 
quire from 60 to 90 per cent, of the 
aggregate horse power of their ma- 
chines. The connected load should be 
estimated where possible from the 
nature of the work done rather than 
from the motor rating, as motors are 
frequently chosen with reserve capac- 
ity. Where there i^ a considerable 
number of motors the maximum load 
is often not more than 40 to 50 per 
cent, of the aggregate rated horse 
power of the motors. Elevator and 
crane motors require transformers of 
100 to 125 per cent, of their rated 
capacity unless there are several mo- 
tors supplied by one unit. This is 
necessary in order to hold up the pres- 
sure in starting. The load of such 
equipments is so intermittent that 
heating is usually not a factor in de- 
termining the size of the transformer. 

Cost of a Single-Phase Line Equipment 


ANYONE who lias gone deeply 
into the subject of railroad elec- 
trification costs cannot fail to be 
impressed with the courage and pro- 
e spirit of the N. Y.. N. H. & 
H. R. R. in venturing on such a big 
experiment as the application of the 
single-phase system to their lir, 

That the system will work, and 
work well when everything is ad- 
justed, is a foregone conclusion, but 
it is interesting, at this time, to con- 
sider what it has cost in order to form 
some idea of the availability of the 
tem for general railroad electrifi- 

While the Xew Haven engineers 
have not, and probably will never 
publish the details of cost of their 
tem, the descriptions which they have 
permitted to be published from time 
to time now enable an engineer with 
estimating experience to form a fairly 
stimate of what the system has 

The estimate given below is based 
on the Xew Haven R. R. descriptions 
which have appeared at various times 
in the technical press, and on unit 
prices now prevailing. The details 
are tabulated at the end of this article, 
and it should be noted that the esti- 




Per mile of Per mile of 
single tra racks 

bridges, intermediate - , everv 300 ft., wgt. 13,000 i 


Steel bridges, anchor; every 2 mi". _ '..000 lbs.. 

Foundations for intermediate bridge, 9 cu. yds. each 

e 34 per mile 

Foundations for anchor bridge, 12 c-j. yds. , I per mile. 

Special foundations 

Trolk No. 0000 B. & S.. 5 - 

Me>;. !0,900ft 

Har.~ apart 

Insulators, two every 30" ft 

Pins and yokes for above 

□ insulators and accessories. 16 every two miles. 
Trolley strain insulators and section breaks. 4 t 

- .:les 

Circuit brt 

Linesmen's materials 

Labor on trolley, messengers, and supports 

for Cont a 

115 tons 

306 j 
12 j 



a i lbs. 







500 . oo 


" - 







Quantity — Unit Price 

Per mile of 
single track 

Per mile of 
four tracks 

No. B. S 10,900 ft 

8 1 lbs. 





$ .18 







Circuit breakers 


rol wire and pipe 


' per section. . . . 




Labor on feeders 




1 b90 





Per mile of 

Per mile of 

four tracks 

Bonds. Xo. O'.XX). B. & S., 32 in.. 

nnina] ex- 


- -• 




mate is probably very low on account 
orf the omission of special work at low 
street bridges and of extra large 
bridge foundations, which occur very 


Per mile of 4-track line. 

Contact System $36,388 

Feeder System 1.890 

Track Bonds 1,800 

Total $40,078 

While the cost of the Xew Haven 
power-station may be readily esti- 
mated (being from $85 to $90 per 
rated kilowatt), it depends principally 
1 'ii the train tonnage to be moved and 
very little on the length of line. It 

is therefore of no interest to classify 

it with the items which depend di- 
rectly on the track mileage, as the 
total would then be inapplicable to 

3 -^ other svstems. 

1 20 

According to the Xew York Cen- 
tral time-table, it is 440 miles from 
Xew York to Buffalo, and the line 
is described therein as four-track. 
Using the above unit price per mile, 

;;;;;;;;; the cost of electrifying the Xew York 
Central to Buffalo would be about 
40,000 by 440 = $17,600,000. exclu- 

20308 sive of power-stations and locomo- 

36.338 tives. This is a very low figure, as no 

- special work is allowed for such as 

would be required going through 

Syracuse in the open streets, etc.. and 

=== ^ = under the hundreds of street and road 

It would appear that the electrifica- 
tion of trunk lines will be retarded 
until some economies are effected in 
contact line cost. 

An item having a commercial a- 
well as humanitarian aspect is the 
danger of the high-tension trolley to 
employees. The loss of life on the Xew 
Haven system from 11, 000- volt 
shocks has been very serious ; the au- 
thor has no exact statistics to date, but 
remembers seven cases during what 
was practically experimental opera- 
tion. The result of this at the present 
time is great delay in executing re- 
pairs, owing to the unwillingness of 
employees to touch the wires without 
elaborate precautions. This contr 
most unfavorably with the 600-volt 
third-rail system, which is easily in 
repair while alive. 


The Richmond and Chesapeake Bay Single-Phase 

R.ailway, Richmond, Va. 


THE accompanying illustrations 
and drawings show the construc- 
tion and the electrical equipment 
of the Richmond and Chesapeake Bay 
Railway, a single-phase electric line 
recently placed in operation in Vir- 
ginia and extending a distance of 
about 15 miles to Ashland from the 
City of Richmond. 

The power for operating this single- 
phase electric railway is supplied by 
the Virginia Passenger and Power 
Company from its Twelfth-Street 
power-house at a pressure of 6600 
volts. The current is transmitted to 
the trolley line at Richmond by under- 
ground cables, and is utilized directly 
at the above pressure without being 
transformed in any way, no substation 
being utilized except for lighting serv- 
ice at Ashland. 

The electrical equipment was in- 
stalled by the General Electrical Com- 
pany, and the cars were built by the 
St. Louis Car Company, the trucks be- 
ing manufactured by the Baldwin 
Locomotive Works. 

The cars are about 54 ft. long over 
all. and 9 ft. 10 in. in width, the bot- 
tom framing of the car consisting of 
yellow pine side sills reinforced by 
steel plates, the intermediate sills be- 
ing 6 inch I-beams. The electrical 
equipment of the cars is of the 
multiple-unit type, with four General 
Electric motors of the single-phase 
railway type, arranged with duplicate 
control. It is stated that the apparatus 
is so arranged that each pair of mo- 
tors and the compensator, as well as 
their contactors, are entirely inde- 
pendent of the second similar equip- 
ment, so that in case of any accident to 
any of the apparatus the other two 
motor equipments may be utilized for 
operating the car. For transferring 
the auxiliary circuits for heating and 
lighting from one compensator to the 
other double throw switches are used. 

It is held that oil-cooled railway 
compensators, designed for 6600 volts 
at 25 cycles, are arranged with taps 
giving from 600 volts down to 113 
volts, and that these are the first Gen- 
eral-Electric equipments put into op- 
eration with two compensators. On 
the auxiliary circuit only are the 600- 
volt taps used, and the cables are in- 
troduced into the tanks, which are 
made of fluted steel through special 
stuffing boxes. 

The cars are provided with two 
trolleys of the pantograph type built 

with steel pans, which are said to be 
better than aluminum or copper, these 
pans being provided with grooves for 
receiving a lubricant. 

The cars have wheels 38 in. in di- 
ameter on axles, d]/ 2 in. at the center 
and 7% ' n - at the gear seats. The 
wheel base is yy 2 ft. and total weight 
of each truck is about eight tons. 

The four motors have a capacity of 
500 h.p., and are of the series repulsion 
type designed for sustaining very 
heavy overloads without injury. Each 
of the motors has a rated capacity of 
125 h.p., but is capable of developing 
far greater power, the overload ca- 
pacity, it is claimed, being due to the 
fact that the whole space in the slots 

E. P. Allis type have been installed, 
each having a capacity of 1000 h.p. 
Tbese engines have a stroke of 3^ ft., 
the high-pressure and low-pressure 
cylinder measuring 20 in. and 40 in. 
in diameter, respectively. The en- 
gines are directly coupled to 750 kw. 
direct-current generators of the Gen- 
eral Electric type. The hydraulic 
turbines are located under the boiler- 
room, the latter being equipped with 
six water-tube boilers of the Babcock 
and Wilcox type, each having a ca- 
pacity of 500 h.p. with a steam pres- 
sure of 145 lb. 

In this power-house of the Virginia 
Passenger and Power Company is 'in- 
stalled the unique generating equip- 


is available for copper, and that there 
is no excessive heating as when the 
leads are of high resistance. 

At the power-house on the James 
River there are steam engines pro- 
vided, as the hydro-electric equip- 
ment cannot be depended upon on ac- 
count of the variable capacity of the 
river, although a dam has been built 
to provide for water storage. 

In order to have sufficient power 
available at all times, five vertical 
tandem compound engines of the 

ment of the Richmond and Chesapeake 
Railway. This machinery included 
two three-phase railway generators of 
750 kw.. each operating at a speed of 
128 rev. per min., and supplying a 
current having a frequency of 25 
periods and a pressure of 6600 volts, 
or 13.200 volts. 

There is a hydraulic turbine pro- 
vided of 1450 h.p. with one of the 
above-mentioned generators directly 
connected at one end, and at the other 
end of the turbine a 750 kw. three- 




April, 1908 

phase generator supplying a 2300-volt 
current at a frequency of 60 cycles. 

This really amounts to a direct- 
connected motor-generator set, the 

be used as a motor when there is not 
enough water to operate the hy- 
draulic turbine. The 1000-h.p. en- 
gines of the vertical type referred to 


electric units being mounted on a 
common base and so arranged that the 
three-phase 60-cycle generator may 

<?0 ! 0" 


above operate direct-current genera- 
tors of 750 kw. each, and the cur- 
rent from these units may be used for 
operating another motor-generator 
set, which is also combined with a hy- 
draulic turbine. 

electric power in case the hydraulic 
turbine cannot be used. 

It is stated that when water-power 
is available for driving the railway 
company's generator the direct cur- 
rent, as well as the alternating-current 
machines, may be utilized for city 
lighting, and the railway company's 
generator may be driven by a direct 
current, or by water-power, or by an 
alternating current of 60 cycles. It 
is stated that one phase of the 25- 
cycle three-phase alternators is used 
to supply the single-phase 6600-volt 
alternating current under the present 
operating conditions. 

The Richmond and Chesapeake Bay 
Trolley line is supplied with a single- 
phase alternating current of 6600 
volts from the Terminal Depot at 
Richmond, it being conducted to this 
point by an underground transmis- 
sion cable in a conduit of vitrified 
earthenware, the total distance being 
about 1^2 miles from the depot to the 

The overhead construction, as seen 
in the accompanying illustrations, is 
of the suspended catenary form, with 
Xo. 0000 grooved copper wire for the 
trolley itself with a seven-strand steel 
messenger cable ^ in. in diameter 
provided, which has a tensile strength 
of 5^ tons. The vertical columns 
are made up of steel pipes three 
inches in diameter with a horizontal 
bar two inches in diameter, malleable 

H) 7 L 0"- *\ 




This outfit consists of a water 
turbine directly coupled to a 25-cycle 
alternator at one of the shafts and a 
direct-current generator of 750-kw. 
capacity at the other end, this dynamo 
being used as a continuous motor for 
operating the 25-cycle generator by 

iron elbows being used. The poles are 
40 ft. in height and the trolley is lo- 
cated 22 ft. from the rail level for the 
overhead construction of the bracket 
type between the Ashland terminus 
and the Richmond viaduct. 

The insulators are 6}i in. in di- 

April, 1908 



ameter and are 4^ in. high, tested 
to a pressure of 50 thousand volts and 
supplied by the Locke Insulator Com- 
pany and the General Electric Com- 
pany. The poles are placed 120 ft. 
apart on the level tangent track, 
while at curves and highway cross- 
ings the spacing is reduced to 60 ft. 

There is one substation at Ashland 
for reducing the pressure from 6600 
volts to 440 volts, and from the low- 
tension transformer terminals leads 
are passed through a phase-splitting 
device, which consists of a resistance 
and a reactance, before being con- 
nected with the induction motor of 
the single-phase alternating-current 

The motor-generator set provided 
consist of this single-phase induction 
motor, above referred to, directlv 
coupled to a single-phase alternator, 
which supplies current to the lighting 
feeders at a pressure of 2300 volts. 

The two step-down transformers, 
above mentioned, are of the oil-cooled 
type of 150-kw. capacity, and they 
supply current to the 150-h.p. motor 
at 440 volts and 25 cycles, while the 
100 kw. single-phase generator of the 
motor-generator set, operating at a 
speed of 720 rev. per min., supplies 
a single-phase alternating current 
having a frequency of 60 cycles, at 
which frequency the current is used 
for lighting the station with arc 
lamps and General Electric tungsten 

The Richmond Terminal Depot is 
built of brick and is provided with 
an electrically operated hoist for rais- 
ing the freight to the station floor 
level. The terminal depot at Ashland, 
containing the lighting substation 

above described, is a wooden building 
with platform extending entirely 
around the structure eight feet in 

There is a viaduct of reinforced 
concrete more than half a mile in 
length constructed to carry the track 

of 14.8 miles. These rails are spiked 
to the ties, which are of white oak 
8 l /> ft. long and 7 by 8 in. in cross- 

It is stated that the maximum curve 
on this single-phase electric railway 
is 412 ft. long and a trifle over 7 


through the suburbs of the City of 
Richmond at a high level varying from 
18 to 70 ft. above the street. The 
spans are from 23 1/3 ft. to 66 ft. 
in length, and expansion joints are 
provided every 200 ft. for arranging 
for the variation in temperature. The 
rails used are of the standard 80-lb. 
type over the whole length of the line 

degrees, while 1 per cent, is the 
maximum grade on this single-phase 
road. It is maintained that the first 
extension considered for this line is 
from Ashland to Tappahannock on the 
Rappahannock, and it is held that the 
construction of a line through this dis- 
trict would be very profitable not only 
for passengers but for freight traffic. 

A SHort Circuit Interrupter 


Long direct-current feeders have 
the serious fault of being liable to 
short circuits, or grounds, which may 
be of sufficient magnitude to cause 
great damage, but not to open circuit 
breakers set high enough to meet the 
regular load demands in heavy trac- 

In order to meet this condition, the 
writer devised the scheme outlined be- 
low whereby a short circuit, or ground, 
is made distinguishable from a heavy 
legitimate load, so that the circuit 
breaker will respond to the former, 
but not to the latter. 

The system consists in incorporating 
in, or attaching to, the feeder cable a 
small insulated wire, which is con- 
nected in circuit with a relay and 
source of current in such a way that 
the relay will operate when this small 
wire is severed at any point, either 
from a short circuit or any other 
cause. The relay is connected to the 

shunt trip coil of the circuit breaker, 
so that the operation of the relay 
actuates the latter and opens the 
circuit breaker. 

This scheme was worked out by the 
writer about two years ago, and has 
been put into use on the direct-current 
feeders of the New York Central Rail- 
road electric system. Among the 
features worked out at the time re- 
ferred to above was the application of 
the system to third rails, and the in- 

corporation in the feeder cables of an 
insulated strand to be used in con- 
nection with the system. The use of 
the system by the New York Central 
Railroad was described by Mr. W. J. 
Wilgus before the American Society 
of Civil Engineers, and had been else- 
where commented upon favorably by 
New York Central officials. 

Premature publication of the sys- 
tem prevented patenting, so that it is 
now available to all who wish to use it. 





(If there were two cables, the return, instead of being grounded, would be the wire on 

the second cable.) 

The Problem of Illumination 



IN most of the fields in which the 
engineering- sciences are applied 
to forwarding the world's comfort 
and to doing the world's work, the 
question of the efficiency of the ap- 
paratus employed and of the installa- 
tion as a whole is one which must re- 
ceive the very serious consideration of 
both the designing engineer and the 
man who pays the bills. In few of 
these applied sciences, however, is 
efficiency of such supreme importance 
as in illumination. In the application, 
for instance, of electrical energy to 
mechanical work through the agency 
of the electrical motor, the primary 
consideration is usually the reliability 
of the apparatus for service whenever 
required, while efficiency, though an 
important, is none the less a secondary 
consideration. In the air-brake or in 
the signal system, reliability of serv- 
ice is of vastly more importance than 
operation at an efficient power con- 
sumption. In illumination, however, 
efficiency of operation is a considera- 
tion so paramount that it may rightly 
be called the problem of illumination.* 
I fancy there are two possible ques- 
tions which may be raised by those 
who hesitate to accept the statement 
that the problem of illumination is es- 
sentially a problem of efficiency. "We 
agree." one party may say, "that the 
issue is one of satisfactory illumina- 
tion at low cost ; but is efficiency of 
operation the all-important factor of 
the cost? Is not the useful life of 
the illuminating apparatus of equal or 
greater importance?" The question 
here raised can easily be settled be- 
vond dispute and settled once for all. 
Those who are most apt to lay great 
importance on the question of useful 
life are those who use the incandescent 
lamp as a source of illumination. In 
Fig. i is shown the relation which the 
renewal cost of the lamp bears to the 
power consumption cost at different 
costs of power per kilowatt-hour. The 
curves apply to the 16-c-p.. 3.1 watts 
per c-p.. no-volt carbon filament 
lamp. Curve A represents the renew- 
al cost expressed as a percentage of 
the total cost of operation, including 
renewal. Curve B represents power- 
consumption cost, likewise expressed 
as a percentage of the total cost. This 
curve sheet shows that, at all costs at 

which it is practicable to generate 
electrical energy, the power-consump- 
tion cost is considerably greater than 
the renewal cost. Within the range 
of cost at which electrical energy is 
commonly available to the user of 
light, 8 cents to 15 cents per kilowatt- 
hour, the renewal cost is an almost 
negligible factor as compared to the 
power-consumption cost — I have chos- 
en the most efficient carbon filament 
lamp to illustrate the case ; had a less 
efficient lamp been taken, the life of 
the lamp would have shown up as a 
still less important factor. 

It is obvious that in the case of gas 
in open burners, and of spirits or oil 
illuminants, the life of the illuminating 
apparatus is a very trivial factor in 
the total cost of operation. In the 
case of the gas mantle, Nernst lamp, 
mercury vapor lamp, or any of the 
various forms of arc lamp which are 








TO ' 

>o i 

to • 

O 6 

9 £ 

O 4 

O J 

to Z 




many another applied science, art and 
efficiency go hand in hand. Not all 
so-called artistic installations are effi- 
cient, but the truly efficient installa- 
tion is almost invariably artistic in the 
best and highest sense of that much 
abused word. And therein lies the 
answer to our critics. Efficiency' 
comes first. We must seek efficiency 
with a single eye and satisfy its can- 
ons. Only by so doing can we lay the 
secure foundations upon which the su- 
perstructure of true art must be built. 
Perhaps the expression of these 
principles in general terms seems ab- 
stract, almost meaningless. Take a 
concrete example. A false sense of 
the artistic, which finds much favor at 
the present time, decrees that the in- 
candescent lamp must not be sus- 
pended vertically from the chandelier, 
but at an angle to the vertical. The 
same artistic sense surrounds the lamp 






/^srCenT of To fa/ Cost Of OperoT/on . 

FIG. I. 

A'paper delivered before the Pittsburg section of 
the •American Institute of Electrical Engineers at 
meeting January 8, 1908. 

in extended commercial use, the re- 
newal cost bears to the power-con- 
sumption cost a relation similar to 
that which we have already found true 
for the incandescent lamp. 

We have seen that the useful life of 
the illuminating apparatus is of minor 
importance in the problem of illumina- 
tion, as compared with efficiency of 
operation. There is yet one other 
party of critics who have a right to be 
heard. These will assert that the ar- 
tistic quality of the installation is fre- 
quently of more importance than its 
efficiency. If artistic merit and effi- 
ciency were antagonistic, or even in- 
dependent principles, these critics 
would be right. We do desire beaut v 
in our surroundings, and we will have 
it, if necessary, at the cost of dollars 
and cents. But in illumination, as in 

with a reflecting glass shade. With 
such an installation, whenever one 
faces the chandelier even partially, 
one is struck in the face with a beam 
of light. The result is not merely un- 
pleasant, but inartistic in the highest 
degree. Had the problem of efficiency 
first been studied it would have been 
found that the most efficient position, 
the vertical, is also the most artistic in 
the illumination results attained. 

I have thought it worth our while 
to discuss at some length the proposi- 
tion implied in the title of this paper. 
The problem of efficiency is the funda- 
mental problem of illumination. The 
laws upon which efficiency of illumi- 
nation depends are to become the fun- 
damental laws of the new-born science 
of illuminating engineering. Our 
natural and only logical procedure, 


April, 1908 



therefore, as students of the new sci- 
ence, is to analyze the various factors 
which determine the efficiency of il- 
lumination and to recognize and clas- 
sify the laws according to which these 
factors act. 


At the outset it is necessary to have 
a clear conception of what is meant in 
illuminating engineering by efficiency 
In the older and better-established en- 
gineering sciences efficiency is regu- 
larly taken to mean the useful energy 
output of the apparatus or installation 
divided by the energy input, the re- 
sult being expressed as a percentage. 
In the practice of these sciences, the 
end sought is the transformation of 
one purely physical form of energy — 
mechanical, electrical, chemical — into 
another purely physical form of en- 
ergy. In illumination, however, the 
end sought is a physiological process 
— sight. The difficulty of measuring 
efficiency of illumination at once be- 
comes apparent. Watts and horse 
power can be reduced to a common 
unit and the one divided by the other. 
But what common unit can be found 
between watts and the sensation called 
clear vision? Does clear vision under 
different illumination conditions rep- 
resent always the same amount of en- 
ergy expended in the physiological 
process? Can, indeed, any satisfac- 
tory unit of "clearness of vision" be 
found, whether energy unit or other- 

It is foreign to the purpose of this 
paper to follow the inquiry into which 
these and kindred questions would 
lead. It is sufficient for present pur- 
poses to recognize that illuminating 
engineering has as yet no unit of effi- 
ciency. Xone the less, the term effi- 
ciency can be used, and the distinc- 
tions of higher and lower efficiency 
can be drawn. That such practice is 
legitimate a moment's consideration 
will make clear. It has already been 
pointed out that in the older engineer- 
ing sciences efficiency varies directly 
as the useful energy output and in- 
versely as the energy in-put. If, now, 
either useful output or energy in-put 
can be brought to a given definite con- 
dition, relative efficiency at that con- 
dition can be expressed in terms of 
the other quantity. In illumination 
there is, for any given installation, a 
fairly definite condition known as 
"good illumination." For any given 
plane of reference, or for any combi- 
nation of such planes, the relation of 
efficiencies of two different schemes of 
illumination will be the inverse rela- 
tion of the energy in-put required to 
produce "good illumination" in each 

"Good illumination" is indeed a 
rough and inexact measure for a sci- 

ence which seeks to be exact. But 
however ill we like it, we will be the 
better engineers for clearly recog- 
nizing that it is at present the only 
measure we have of useful energy out- 
put — of the energy which is active in 
stimulating the optic nerve and pro- 
ducing visual perception. 

This, then, is the situation in which 
the illuminating engineer finds him- 
self. His problem is fundamentally a 
problem of efficiency — the problem of 
using the energy at his disposal so as 
to accomplish the greatest amount of 
useful work measured in terms of 
visual perception. He has, however, 
as yet no unit of efficiency. He can- 
not, therefore, measure efficiencies in 
absolute terms. He can, however. 
compare the relative efficiencies in in- 
verse terms of energy in-put required 
to produce a certain condition of visual 
perception popularly termed "good il- 
lumination." He can, moreover, de- 
termine the laws on which efficiency 
depends and apply those laws so as to 
produce highly satisfactory practical 
results, even though he cannot express 
in absolute terms the results attained. 


When the term "efficiency of il- 
lumination" is considered closely, it is 
seen that in this term is included the 
combined effect' of three different 
kinds of efficiencies. First, there is 
the efficiency of the light source, by 
which is meant the efficiency with 
which chemical or electrical energy is 
transformed into light energy. Sec- 
ond, there is the efficiency of light dis- 
tribution, by which is meant the rela- 
tion between total light energy gen- 
erated and light energy useful in pro- 
ducing desired conditions for visual 
perception. Third, there is the effi- 
ciency of visual perception, this being - 
the efficiency with which the eye re- 
ceives light energy and transforms it 
into visual perception. Xone of these 
factors is trivial — each is of tre- 
mendous importance in determining 
the resultant efficiency of illumination. 
It has been the error of the past to lay 
all the emphasis on the efficiency of 
the light source. With the develop- 
ment of illuminating engineering, and 
particularly as a result of the admir- 
able work of Mr. Van Rensselaer 
Lansingh, there is coming to be a gen- 
eral realization of the importance of 
light distribution. It remains for the 
future to recognize that efficiency of 
visual perception is a factor which can- 
n< it be neglected. 

Our problem lies clear before us. 
We seek efficient illumination. To 
obtain it we must have efficiency of 
visual perception, of light distribu- 
tion, of light source — all three. What, 
then, are the conditions upon which 

each of these three kinds of efficiencies 
depend ? 


Efficiency of visual perception de- 
pends upon three conditions. These 
are (a) the intrinsic brilliancy of the 
light source and of the surrounding 
light-reflecting objects; (b) the color 
of the light: and (c) the intensity and 
steadiness of the light. Each of these 
conditions will be briefly considered in 

(a) Intrinsic brilliancy of light 
source. It is a well-known fact that 
the eye adjusts itself to various de- 
grees of light intensity by the auto- 
matic expansion or contraction of the 
pupil — that opening in the iris dia- 
phram through which light is admitted 
to the eye. Now, the light which is 
active in causing a greater or less 
contraction of the pupil is not merely 
the light which comes from the center 
of the field of vision, but the light 
which comes from the entire field of 
vision. The light which is active in 
causing visual perception, however, 
comes under normal conditions en- 
tirely from the central portion of the 
field of vision. The same amount of 
light, therefore, falling upon and re- 
flected from the given visualized ob- 
ject may produce very different de- 
grees of illumination due to changes 
in the size of the pupil, such differ- 
ences resulting from differences in the 
intrinsic brilliancy of the outlying 
portions of the field of vision. 

A concrete example will make clear- 
er. One sits reading, let us say, in a 
room with- dark-colored walls, the only 
source of light being above and be- 
hind the reader, entirely outside of 
his field of vision. Photometric meas- 
urements of light intensity on the 
printed page give results of two 
foot-candles. The reader calls it good 
illumination. Xow we will introduce 
a second light source, say a brilliantly 
incandescent W'elsbach mantle or a 
tungsten lamp, into the reader's field 
of vision, screening the printed page 
so that it gets no light from the new 
light source. Gradually we will bring 
the new light source more arid more 
to the center of the reader's field of 
vision. Just so gradually the pupil of 
the reader's eye contracts, admitting t" 
the retina less and less light from the 
printed page. When the new light 
source, still screened from the printed 
page, has been brought near the cen- 
ter of the reader's field of vision, pho- 
tometric measurements are again 
made and the results again announced 
to be two foot-candles. "I know 
nothing about your loot-candles." is 
the reader's impatient reply, "hut I 
eall it might)- poor illumination." 

White or very light colored walls 
and ceilings, if brightly lighted, may 



April, 1908 

have an effect in contracting the pu- 
pil similar to a bright light source in 
the field of vision. 

The competent illuminating engi- 
neer recognizes the serious decrease 
in efficiency which results from bril- 
liant light sources or brilliantly lighted 
white walls, and he embodies that 
recognition in the following concrete 
rules of practice. 

Whenever a brilliant light source is 
so placed that it may come within the 
field of vision, reduce to a low value 
the intrinsic brilliancy of that light 
source by a diffusing sphere, bowl, 
stalactite or bell-shaped shade. Never 
use a bare incandescent lamp, nor the 
Welsbach mantle with no shade other 
than the clear glass chimney or mica 

Do not illuminate light-colored 
walls or ceilings too brilliantly. 

This last rule, and the physiolog- 
ical conditions which justify it, are 
ignored by those who recommend in- 
stallations of what is usually called 

A study of the effect produced on 
the efficiency of illumination by un- 
shaded light sources and by bril- 
liantly lighted white walls is recom- 
mended to those who seek to apply 
the foot-candle as an adequate meas- 
ure of illumination. 

(b) Color of light used. The ef- 
fect on efficiency of visual perception 
arising from the color of the light is 
a subject too involved and at the pres- 
ent time too little understood to war- 
rant discussion. It is sufficient to say 
that for the same light intensity as 
measured in foot-candles, lights of 
different color give appreciably dif- 
ferent illumination values. For il- 
lumination where objects of a great 
variety of color are to be viewed, the 
best light is probably light of the qual- 
ity of summer daylight — that is, light 
containing all wave lengths, but hav- 
ing a slightly undue preponderance of 
the green rays. For illumination of 
black and white effects, as drafting- 
room illumination, it is an open ques- 
tion whether modified white light, 
such as just described, or green light 
as of the mercury-paper lamp, is most 

(c) Intensity and steadiness of the 
light. It is a common fallacy to as- 
sume that the more light, the better il- 
lumination. As a matter of fact, for 
any given integral of light intensity in 
the outer portions of the field of vision 
there is a corresponding definite in- 
tensity of light for the central portion 
of the field of vision which will give 
best conditions for visual perception. 
Greater intensities than this will pro- 
duce gradual or rapid fatigue of the 
eye, resulting in less clear vision. A 
flickering, unsteady light also pro- 

duces rapid fatigue of the eye func- 
tions. For this reason the old style 
open gas burner should never be used 
unless protected from air currents by 
a chimney. Burners employing in- 
candescent mantles are free from this 

It is here intended to point out only 
the relations which intensity and 
steadiness of light bear toward effi- 
ciency of visual perception. In addi- 
tion to this effect on efficiency of vis- 
ual perception, and separate from it, 
is the permanently injurious effect on 
the eye which unsteadiness and too 
great an intensity of light may have. 

To recapitulate, increased efficiency 
of visual perception may be obtained 
by detailed application of the follow- 
ing general rules. Reduce to a low 
value the intrinsic brilliancy of the 
light source on all sides exposed to the 
eye ; avoid intensity of light on light- 
colored walls or ceilings ; use light 
of correct color value ; avoid unsteady 
light, and avoid excessive intensity of 
light on surfaces which are constantly 
or frequently objects of visual per- 


The sect >nd kind of efficiency with 
which the illuminating engineer has to 
deal, efficiency of light distribution. 
depends upon two important factors: 
(a) the distribution of light which 
emanates from the illuminating unit 
and (b) the size of the unit and the 
location of centers of light distribu- 

Light is a very unstable form of 
energy. It may be generated only 
with poor efficiency, and, once ob- 
tained, it slips back under ordinary 
conditions almost immediately into 
other forms of energy. Absorption is 
the chief process by which light en- 
ergy is lost as such. Excluding mir- 
rors and surfaces especially prepared 
for reflecting purposes, when light 
falls upon any of the surfaces with 
which we are commonly surrounded, a 
very considerable percentage is ab- 
sorbed, while the rest, usually the 
much smaller part, is reflected. To 
give average figures, the percentage of 
light which is reflected from different- 
ly colored papers is as follow^ : 

Table i 

White paper 80% 

Orange paper *>0% 

Yellow paper ^®% 

Light pink paper 35% 

Light blue paper 25% 

Emerald green paper lS r r 

Dark brown paper 10% 

These figures make it obvious that, 
if light is to be used efficiently, it 
must not undergo many reflections, 
losing in each as it does under average 
conditions, say 70 per cent, in intens- 
ity. Indeed, ideal efficiency of distri- 
bution is obtained only when all of 

the rays emitted by the illuminating 
unit proceed directly and in proper 
proportion to the various surfaces to 
be illuminated, whence they are re- 
flected into the receiving eye. In 
proper proportion, it is said. For sup- 
pose the illuminating unit emits in one 
direction more rays than required to 
give the proper illumination of the 
surfaces A, B, C on which they fall. 
A large number of these additional 
and unnecessary rays will be absorbed, 
which means that that much light is 
wasted, thrown away. Such of these 
rays as are not absorbed will be dif- 
fusely reflected. Xow these diffuselv 
reflected rays will decrease rather than 
increase the effectiveness of illumina- 
tion of the surfaces A, B, C, for with- 
out them we had the correct degree 
of illumination. Most of these dif- 
fusely reflected rays, however, will 
not enter the eye, but will fall on the 
ether surfaces, D, E, F. Here there 
will be another absorption. The few 
rays that remain may, it is true, be 
useful in the illumination of the sur- 
faces D, E, F, but at what a sacrifice 
of efficiency is such illumination ob- 
tained ? 

The facts here adduced are so well 
known that detailed discussion of 
them seems trite. Nevertheless they 
are largely ignored in the vast ma- 
jority of existing installations. There 
are few present installations in which 
the power consumption required to 
operate cannot be halved, or the effi- 
ciency of illumination doubled, by util- 
ization of the light now needlessly 
wasted through absorption. Here, 
then, is a factor just as important for 
efficiency of illumination and just as 
deserving of attention as the recent 
developments in high-efficiency incan- 
descent lamps. 

It has been stated that the ideal effi- 
ciency of distribution is obtained only 
when all the rays emitted by the il- 
luminating unit proceed directly and 
in proper proportion to the various 
surfaces to be illuminated. When at- 
tempt is made to realize this ideal, an 
apparently serious difficulty is en- 
countered. None of the present light 
sources give of themselves such light 
distribution as to fulfil even approx- 
imately the condition just stated. The 
Nernst lamp and the inverted gas 
mantle give better distribution than 
any of the other largely used light 
sources, but even the distribution of 
these is unsuitable when high effi- 
ciency of illumination is sought. 
Fortunately the distribution of the 
bare light source can be greatly modi- 
fied and highly efficient distribution 
obtained by the use of reflecting or 
refracting shades or globes. Of these 
various means for modifying and 
making efficient the distribution, 

April, 1908 



prismatic glassware stands away and 
above all others as best accomplish- 
ing the desired results. Indeed, where 
strong concentration is wanted in one 
direction and at the same time a small 
amount of light in all other directions 
— a frequently desired form of distri- 
bution — prismatic glassware is the 
only means at present known which 
will accomplish this result. For broad 
downward distribution of light, satis- 
factory results can be obtained by the 
use of suitably shaped opal or green- 
enameled glassware. The etched glass 
shades so frequently used are ineffi- 
cient and unsuited for properly modi- 
fying the distribution. Where large 
areas are to be illuminated from a 
few light sources, excellent distribu- 
tion can be obtained by the use of 
sand-blasted or opal glass globe with 
flat conical reflector of green-enameled 
or opal glass mounted immediately 
above the globe. 

The application of suitable glass- 
ware to give proper distribution re- 
sults is a problem which should not 
be attempted by engineer or layman 
unless he be especially trained and 
fully qualified as an illuminating engi- 
neer. Xc educated man. much less a 
scientifically trained man, would think 
of going to an optician's store and se- 
lecting a pair of eye-glasses because 
the curvature of the lens looks suitable 
to him. or because the mounting was 
artistic. Yet procedure no whit less 
ridiculous is the common occurrence 
in illumination problems. The writer 
has known trained engineers, men in 
some cases of unusual ability, to in- 
stall prismatic glassware without even 
knowing the distribution effects given 
by the particular type in question. 
Sharp downward concentration, per- 
haps, was desired ; vet the reflectors 
were installed in utter ignorance of 
whether they threw maximum light 
downward or maximum light laterally 
with minimum light downward. And 
when poor illumination results were ob- 
tained, as was to be expected, the con- 
clusion was drawn that prismatic 
glassware was a greatly overrated 
product. It is a fact too little appre- 
ciated that the problem of obtaining 
efficient distribution is a complex, 
technical problem, requiring for its so- 
lution highly specialized knowledge of 
the physics of light and of the ma- 
terials of illumination, and a broad ex- 
perience with illumination installa- 

The foregoing may profitably be 

summarized and emphasized as fol- 
lows : 

First, a correct distribution of light 
about the illuminating unit is an indis- 
pensable condition to efficient illumina- 
tion and low cost of operation. 

Second, no light source now on the 

market gives of itself correct distribu- 
tion. Hence suitable glassware to 
modify and correct the distribution 
should invariably be used. 

Third, the problem of correct dis- 
tribution and proper means of obtain- 
ing it is a highly technical problem to 
he solved only by the competent illu- 
minating engineer. 

The size of the illuminating unit and 
the location of centers of light distri- 
bution is a second important factor in 
efficiency of distribution. Facli instal- 
lation is so much a particular problem 
that it is difficult to lay down any uni- 
versal laws expressing 1 the relation of 
these factors to efficiency of distribu- 
tion. The following rules, however, 
will be found to have a general appli- 

The center of light distribution 
should be located over each point at 
which relatively high intensity of light 
is desired. The need of additional 
centers and the location of such varies 
with the particular problem. 

Avoid too many centers of distribu- 
tion. Such design is not only costly to 
install, but less efficient in operation. 

be obtained by means independent of 

the wall brackets. 

In applying the above rule to resi- 
dence lighting, and observing the 
principles already stated that tor) much 
light is just as much to be avoided as 
too little light, it will be found that 
small light sources are required. In 
residence lighting, eases are very un- 
usual where light sources of mean 
spherical candle-power greater than 
20 c-p. can be used without consider- 
able sacrifice in efficiency. A light 
source of mean spherical candle-power 
from eight candle-power to 10 c-p. has 
largest application in residence light- 
ing. It is in this feature that the in- 
candescent lamp has its greatest ad- 
vantage over other illuminants : it is 
readily obtainable in whatever size it 
is most efficient for the particular in- 


The efficiency of the light source is 
the third kind of efficiency with which 
the illuminating engineer has to deal. 
Efficiency of light source depends pri- 
marily upon a single factor— namely. 

f £#£/?6V I 3700' C 


//Vf/fff - /?£0. 

FIG. 2. 

More artistic results and better effi- 
ciency are obtained when the light 
comes from clearly marked sources. 
It should be noted, however, that the 
application of the previous rule will 
sometimes require a large number of 
centers of distribution ; for instance, 
in drafting-room illumination. In 
such cases the previous rule is para- 

Several smaller illuminating units 
adjacently located at the same center 
of distribution give more efficient ef- 
fects than a single larger unit. The 
most suitable number of units for one 
center of distribution is usually three, 
four or five. 

Do not locate an illuminating unit 
on wall bracket or closely adjacent to 
a wall. If wall brackets are insisted 
upon on account of their supposed ar- 
tistic effect, a small light source, sur- 
rounded by a relatively large diffusing 
gli >be. should be used. By this arrange- 
ment the so-called artistic effect is at- 
tained without introducing any objec- 
tionable factors into the illumination. 
The actual illumination will, of course, 

the temperature to which the incan- 
descent light-giving body is raised. 

A clear idea of the relation of tem- 
perature of incandescent body to ef- 
ficiency of light generation may be 
obtained by reference to Fig. 2. The 
curves here shown are copied from an 
article by Mr. E. Percival Lewis in the 
California Journal of Technology of 
April, 1907. The curves graphically 
show the various steps in the phenom- 
enon of light generation. When suf- 
ficient heat is imparted to a body to 
raise it to a low temperature, say 
ioo° C. the body radiates energy in 
the form of long period ether waves. 
These waves are capable oi stimula- 
ting certain nerves of the skin, and 
from this we have come to call them 
heat waves. These waves, it should 
be noted, have not the power to stimu- 
late the optic nerve. I f. now. the 
temperature of the body be raised to 
6oo° C. a larger amount of energy is 
radiated, and the radiations are over 
a broader range of wave-lengths. 
Some of the shorter of the waves now 
generated are capable of stimulating 



April, 1908 

the optic nerve. These waves we call 
light. Again raising the body to a 
higher temperature a still large 
amount of energy is radiated over a 
still broader range of wave-lengths. 
The shorter of the waves now pro- 
duced are incapable of stimulating 
either optic nerve or nerves of the 
skin ; but these very short waves are 
especially active in producing chemical 
changes, and hence they are some- 
times called chemical waves. It will 
be noted that as the temperature rises 
and the total amount of radiating en- 
ergy increases, the increase is chiefly 
in the shorter wave-lengths. Hence 
the proportion of light waves to total 
wave energy, or, in a word, the effi- 
ciency increases as the temperature in- 

The curves apply equally to any in- 
candescent solid whatever its chem- 
ical constitution. They therefore ap- 
ply to all light sources in practical 
commercial use except the mercury 
vapor and other vapor lamps. 

Every body or material has, how- 
ever, its own fairly well-marked limit- 
ing temperature beyond which, if it 
is raised, the structure of the body 
rapidly deteriorates. Carbon, for in- 
stance, when heated in vacuum, 
throws off minute carbon particles 
very rapidly at temperatures above 
1800 C. This, therefore, is the tem- 
perature which limits the efficiency of 
the carbon filament incandescent lamp. 

Fortunately, in the case of both gas 
and incandescent lamp light sources, 
materials have been found which will 
withstand high temperatures without 
rapid deterioration. The gas mantle 
is too well known to need comment, 
but at its introduction it marked a 
tremendous advance in efficiency of 
light sources. Now a still greater ad- 
vance has been made by the develop- 
ment of methods of preparation of 
various metals, of which tungsten 
stands first, for use as the filament of 
the incandescent lamp. Tungsten may 
be raised in vacuum to a temperature 
of at least 2300 C. without the occur- 
rence of rapid deterioration. This in- 
crease in temperature from 1800 C. 
to 2300 C. represents an increase in 
efficiency of two and one-half times. 

A short digression from the subject 
proper of this paper is, I trust, per- 
missible, in order to bring out certain 
salient features of the new high-effi- 
ciency tungsten lamp. 

Fig. 3 shows the relative light dis- 
tribution of carbon filament and tung- 
sten filament lamps when burning at 
equal, wattage. The carbon filament 
lamp is burning at 3.1 watts per mean 
horizontal candle, the tungsten fila- 
ment lamp at 1.25 watts per mean 
horizontal candle. 

Table II shows the comparative effi- 
ciency of the tungsten lamp among 
other electric illuminants. This table, 
with the exception of the value for the 
tungsten lamps, is taken from Cravath 
and Lansingh's Practical Illumination. 
Values for the tungsten lamp were ob- 
tained from tests made in the labora- 
tories of the Westinghouse Lamp 
Company, and are correct for the type 
of lamp tested to within -f- or — 3 
per cent. 

The brittleness of tungsten is so ex- 
treme that it seems at present very 
improbable that lamps of lower 
candle-power than 25 mean horizontal 
can ever be commercially made for 
operation on no- volt circuits. At the 
present state of development, the 
commercial practicability of a lamp of 
lower candle-power than 40 has not 

lamp can be produced with filament 
strong enough to withstand shipment 
with small percentage of breakage. 
The tungsten filament lamp, on ac- 

All values involving candle-power are expressed in 
terms of mean spherical candles. 





of Light 

Kind of Lamp 



per Kilo- 






Common 56 watt carbon 

filament incandescent 

lamp rated at 3.5 w. p. 

c. ; 16 horizontal c. p.. . 


4 24 


Common 50 watt carbon 

filament incandescent 

lamp rated at 3.1 w. p. 

c; 16 horizontal c. p... 




3-glower 264 watt Nernst 




High-efficiency Gem 1 25 

watt graphitized carbon 

filament lamp of 50 hor- 

izontal c. p 




44 watt tantalum lamp 22 

rated horizontal c. p. . . 


2 75 


Direct current 5.1 ampere 

enclosed arc on 110 volt 

circuit, 0.5 inch carbons 




Alternating ■» current en- 

closed 5.7 amp. arc tak- 

ing 388 watts on 110 

volt circuit 05 inch car- 



2 55 


Tungsten 60 watt 1.25 w. 

p. c. 110 v. lamp 




Luminous 8 amp. arc, 440 

watts, 2 in series on 110 




FIG. 3. 

been fully proven. The 40 c-p. lamp, 
or even the 25, is too large a unit for 
most residence lighting. At present 
writing it seems very probable that for 
residence lighting some plan of cir- 
cuits will have to be adopted so as to 
give available voltages of 25 volts or 
50 volts across the lamp terminals. At 
25 volts a 10 c-p. (mean horizontal) 

count of the positive temperature co- 
efficient of tungsten, is very much less 
injured by variations in voltage than 
the carbon filament lamp. Double 
voltage may be thrown across the 
terminals of the tungsten filament 
lamp without causing immediate burn- 
out, the average lamp burning at such 
double voltage for over an hour. 

Downward Versus Horizontal Illumination 

THE author of the article entitled 
"Downward Illumination" in the 
March issue of The Electrical 
Age has taken a wrong, though a popu- 
lar, point of view in considering the 
advantages of the "downward light" 
lamps that have been put on the 
market. He shows curves which in- 
dicate the well-known fact that the 
flattened coil and spiral filament lamps 
give a higher tip candle-power than 
do the ordinary oval anchored carbon 
or the tantalum filament lamps. He 
then endeavors to show why these 
"downward light" lamps should re- 
place the ordinary types when a 
strong illumination is needed directly 
under the lamp, leaving almost out of 
consideration the use of proper re- 

The "downward light" lamps get 
their downward distribution on ac- 
count of having a large part of the 
filament arranged approximately hori- 
zontal. They can be designated, then, 
as "horizontal filament" lamps as op- 
posed to the ordinary, or we may say, 
the "vertical filament" lamps. Let us 
consider the distribution curve of one 
of these horizontal filament lamps 
( Fig. i ) . This is the curve for a 
spiral filament. It shows the charac- 

downward as upward. Most of the 
light radiated upward from the hori- 
zontal portions of the filament is lost 
in the base by absorption. If it were 
not for the presence of the stem and 
the base of the lamp, the distribution 
curve would be almost symmetrical on 
both sides of the horizontal axis, since 
the filament is approximately sym- 
metrical. Let us now draw a dotted 
curve above the horizontal to corre- 
spond with that half of the distribu- 
tion curve below, thus making it sym- 
metrical. Now from the area (shown 
shaded) between this dotted curve 
and the curve of distribution above 
the horizontal, we get an indication 
of the amount of light that is lost in 

FIG. I. 

teristic high candle-power in the 
downward direction, obtained by the 
approximately horizontal arrangement 
of the filament. Since the filament 
must emit light equally in all direc- 
tions normal to its surface, there must 
be the same intensity of light emitted 

FIG. 2. 

the base of the lamp when this shape 
of filament is used. The area is large, 
indicating a large loss. 

The curve in Fig. 2 represents the 
distribution about the tantalum lamp, 
one of the most pronounced types of 
the vertical filament lamps. Here 
there is a low tip candle-power indi- 
cating that but little light is thrown 
downward and upward. When the 
dotted curve is drawn as it was with 
the horizontal filament lamp the area 
enclosed . is very small, indicating a 
very small loss of light in the base. 
Hence, for a given watt per candle 
consumption the bare vertical filament 
lamp would be more economical than 
the bare horizontal filament lamp. 
The horizontal filament lamp has, 
however, the advantage in the fact 
that its distribution is more often de- 

sirable than that of the vertical fila- 
ment lamp. 

A bare incandescent electric lamp 
having a filament of whatever shape is 
very seldom a proper lighting unit. 
Hence, to get a proper distribution, 
both sorts of lamps will often require 
the use of reflectors. Near the base 
of a lamp all light that has its course 
changed downward by reflectors must 
strike the reflector at a large angle. 
This, of course, is a condition that 
favors absorption losses in the re- 
flector. With the vertical filament 
lamp the radiation in these unfavor- 
able directions is less than in the hori- 
zontal filament lamps, and is superior 
for that reason when reflectors are 

The case of the horizontal filament 
lamp, then, may be stated as follows: 
(i) It gives in some instances a dis- 
tribution approximately that desired; 

(2) this distribution is obtained at 
the cost of a considerable loss of light 
in the base; (3) if it is desired to 
change its distribution by means of a 
reflector, light of fairly high intensity 
must strike the reflector at large an- 
gles, which is a condition favoring 
absorption losses. In regard to the 
vertical filament lamp it is observed: 
(1) It gives in few instances the dis- 
tribution that is desired; (2) there is 
little loss of light in the base, because 
little of the filament is horizontal ; 

(3) when the distribution must be 
changed by a reflector there is little 
light that strikes the reflector at large 
angles and then it is of low intensity, 
thus giving low absorption losses in 
the reflector. When one considers the 
fact that the best reflectors are inex- 
pensive, that they can be obtained to 
give any desired distribution and that 
they greatly enhance the beauty of the 
light a vertical filament lamp equipped 
with a reflector seemed a much better 
proposition than a horizontal filament 
light of the same watt per candle con- 
sumption either with or without a 

T. H. Amrine. 

This communication was received 
too late in the month for us to make 
a reply. Next month we shall publish 
a full answer to Mr. Amrine's criti- 
cism. — ED. 


Electric Locomotives — Concluded 

H. L. KIRttER 

THE electric locomotive is not only 
more economical than the steam 
locomotive, but it is simpler. It 
is equally as reliable, and gives a 
smokeless service. These are impor- 
tant factors in the question of substitu- 
tion of the electric for the steam serv- 
ice. The advisability of substitution, 
however, does not directly concern us 
here. It is a question that the manage- 
ment has already settled, so far as the 
St. Clair Tunnel is concerned, or I 
would not be talking to you this eve- 
ning. The interesting fact is that al- 
ternating-current locomotives are sub- 
stituted. Our concern, then, is to get 
a working notion of alternating elec- 
tric currents. 


I pointed out that the speed charac- 
teristics of the series motor, the eco- 
nomic advantage of electric traction 
and the desirability of smokeless serv- 
ice are commercializing the electric 
locomotive. There are no indications, 
however, that the direct-current loco- 
motive will eliminate the steam loco- 
motive from main-line work. The 
trolley-voltage limit settles that point. 
There are but few traction systems 
that operate direct-current motors on 
a voltage greater than 650 volts. 
There have been many improvements 
in motor design within the last ten 
years, but the voltage limit seems 
to remain about 650 volts. Heavy 
power, we known, cannot be trans- 
mitted long distances economically 
with such a direct-current trolley 
voltage as this, even if special means 
of generation and transformation are 
resorted to. I also pointed out that 
there is a considerable waste of energy 
every time the direct-current motor 
starts. Then there is the electrolytic 
effect of direct current. You know 
that direct current has a dissolving ef- 
fect on metal pipes at points where the 
current flows from the pipes into moist 
earth. Minute particles of the metal 
seem to travel along with the current 
at these points. The tracks usually 
constitute an important part of the 
return circuit. It is not surprising, 
then, that we find in the large urban 
systems that some of the current 
strays from the proper returns and 
finds its way to gas pipes and water 
pipes and thence into earth, as the cur- 
rent makes its way back to the station. 
The damage these stray direct cur- 
rents do to these pipes is sometimes 


The natural solution of the electrol- 
ysis problem is, of course, to neutral- 
ize this dissolving tendency. Suppose 
our current instead of being contin- 
uous is a rapidly alternating one. We 
can readily imagine that the minute 
particle of metal that is caught up by 
the current and carried to earth is im- 
mediately returned to the pipe a frac- 
tion of a second later when the cur- 
rent reverses and Hows from the earth 
to the pipe. When a 25-cycle current 
is used the same particle will travel 
back and forth 25 times per second. 
Had continuous current been used, 
why 25 particles of metal would have 
been permanently transferred each 
second from the pipe to earth. In 
fact, we know from experiment that 
the electrolytic effect of alternating 
current on iron pipes is negligible. 
Consequently, if we can make our se- 
ries motor work satisfactorily on al- 
ternating current we can, by the use 
of alternating current, eliminate the 
electrolysis of gas and water pipes. 

But will the series motor work on 
alternating current? A glance at the 
diagram settles this, for we know that 
if we reverse the field current alone 
we reverse the direction of rotation ; 
also, that if we reverse the armature 
current alone we reverse the direction 
of rotation, but if we reverse the field 
current and armature current simul- 
taneously, why the direction of rota- 
tion is unchanged. Reversing the line 
current must reverse the field and 
armature current at the same time, so 
does not affect the rotation (refer Fig. 
25). So far as the series motor is 
concerned, then, we can use either 
direct or alternating current. 



FIG. 25. 

You know, of course, what alterna- 
ting current is — that 25-cycle current, 
for instance, flows alternatingly out 
through the trolley and back through 
the rail, then out through the rail and 
back through the trolley, 25 times in 
each direction per second. If it flows 
alternatingly 60 times in each direc- 
tion per second it is 60-cycle current. 
You also know that, by the use of in- 
duction coils, alternating current can 

be raised to a high voltage and like- 
wise, by the use of induction coils, it 
can be reduced to low voltage again. 
This induction feature makes possible 
the economical transmission of alter- 
nating current over long distances. 
Moreover, we shall see presently that 
the same induction feature enables the 
alternating-current series motor to 
eliminate the losses incident to start- 
ing the direct-current motors. 

Alternating current is the kind that 
Faraday discovered. It is the kind 
you get by moving a wire back and 
forth across a magnetic field. You 
will recall that a commutator had to 
be devised to make the current flow 
continuously in the same direction in 
the line to which the armature is con- 
nected. Now the simple alternating 
current, or single-phase current, as it 
is called, seems to solve the electrol- 
ysis problem, extend indefinitely the 
range of economical transmission and 
minimize the starting losses. These 
three facts mean that the alternating- 
current series motor has a vast field of 
usefulness before it. However, this 
simple alternating current, which is so 
easily produced and which so easily 
induces other alternating currents, has 
been a long time in coming to the 
front as a suitable current for traction 
purposes. It has been split up into 
two-phase and three-phase currents. 
The three-phase current is almost uni- 
versally used for long-distance trans- 
mission and for constant-speed motor 
work. There has been no difficulty in 
producing alternating currents, single- 
phase or polyphase, nor has there been 
much difficulty in producing a con- 
stant-speed polyphase motor. They 
have all been in use for a long time. 
The problem has been to develop a 
variable speed motor that will start on 
a single-phase current. The solution 
of this problem involved the proper 
handling of the induction effects that 
are always possible with alternating 
current. Years of experience in op- 
eration and design have finally solved 
the problem, and we now have a 
single-phase series motor that can 
compete with the direct-current mo- 
tor. The single-phase series motor is 
the kind you will have on the tunnel 
locomotives. So far as we are con- 
cerned, we can consider your motors 
as direct-current series motors. What 
I want to especially direct your atten- 
tion to is the induction property of the 
alternating current. It is easily under- 
stood in the terms of Faraday's dis- 

April, 1908 




I just stated that it is an easy matter 
to produce alternating current. We 
can get it from our direct-current 
armature by tapping the winding at 
two diametrically opposite points, and 
connecting tap Xo. i to an insulated 
slip ring on the armature shaft, which 
slip ring is connected to the trolley 
wire by an insulated brush, and then 
connecting tap Xo. 2 to another sim- 
ilar slip ring which is connected to rail 
in the same way as Xo. I is to trolley 
(see Fig. 26). Assume that the di- 
rection of the field and the direction of 
rotation is to be the same as before. 
Consider what is going on when tap 
No. 1 is in the top position. We see 
that current is flowing up both halves 
of the armature winding and out 
through tap Xo. 1 to the trolley, 
thence down through the motor and 
back through the rail to tap Xo. 2. 
Xow a half revolution later tap Xo. 2 
is on top and we see that the current 
is flowing up through both halves of 
the armature winding and out through 
tap Xo. 2 to rail, thence up through 
the motor and back through the trolley 
to tap Xo. 1 (see Fig. 27). Another 


FIG. 26. 

half revolution brings tap Xo. 1 on 
top again, and the armature currents, 
which are always flowing up, flow out 
through the trolley again. So the cur- 
rent goes on reversing, making a com- 
plete cycle for every revolution of our 
armature. The same armature could 
supply direct current through the 
commutator to one circuit at the same 
time that it supplies alternating 
through the slip rings to another cir- 
cuit. It could also run as a direct- 
current motor and supply alternating 
current through the slip ring, or run 
as an alternating-current motor and 
supply direct through the commuta- 
tor. You know that this latter ar- 
rangement is used in street railway 
work for changing alternating current 
into direct current. Such a machine 
is called a rotary converter. When 
acting as a rotary converter, however, 
the machine is usually supplied with 
three-phase current. I will remark in 
passing that we can get three-phase 
current by tapping our armature 

winding at three equidistant points. 
Three slip rings and three line wires 
are required (see Fig. 28). We see 
that when tap Xo. 1 is on top the cur- 
rent flows out wire Xo. 1 and back 
Xo. 2 and Xo. 3. When Xo. 2 is on 
top the current flows out Xo. 2 and 
back Xo. 3 and Xo. 1, and when X T o. 
3 is on top the flow is out Xo. 3 and 
back Xo. 2 and Xo. 1. However, 
when single-phase motors are used 
direct current is not required, so the 
rotary converter is eliminated, and, 
consequently, the three-phase current 
is unnecessary. 

So much for the alternator ; but 
how about the induction coil, or trans- 
former, as it is called? Well, it is 
based on Faraday's discovery that cut- 
ting lines of force induces a voltage, 
and on the further fact that voltage 
varies with the rate of cutting. Sup- 
pose we wind 1000 turns of insulated 
wire on an iron ring and connect the 
ends to an alternating-current circuit 
(see Fig. 29). When the current is 
flawing into the top end of the coil the 
direction of the magnetism in the 
ring is down through the coil. When 
the current is flowing into the bottom 
end of the coil the direction of the 
magnetism in the ring is up through 
the coil. When the current is zero 
there is no magnetism in the ring. 
Consider loop Xo. 1 when the current 
begins to flow. The incipient rings of 
magnetism dilate until they lie com- 
pletely in the iron circuit. As they 
dilate they cut across all the loops. 
Likewise, the magnetism due to the 
current in coil Xo. 2 cuts across 999 
loops. So does the magnetism of 
loop Xo. 3. So does the magnetism 
of every one of the 1000 loops. We 
see, then, that each loop is cut by the 
magnetism due to 999 loops. Xow 
as the current dies away the rings of 
magnetism contract and disappear. 

her of lines is such and the frequency 
is such that each loop generates one 
volt. The combined voltage of the 
1000 loops then is 1000 volts. This 
voltage set up in the coil by the pas- 
sage of alternating current through 
the coil is called the voltage of self- 

An examination as to the direction 
of the voltage of self-induction shows 
that it is opposed to the current that 
produces it. We found a similar set 
of conditions in the motor armature. 
There the induced voltage opposes the 
current that rotates the armature. In 
order that a current may flow, the line 
voltage must be greater than the in- 
duced voltage. We see, then, that the 
voltage supplied by the alternator 
must be slightly greater than 1000 
volts in order that the alternator may 
drive a current against the 1000 volts 

in,. 28. 

of self-induction set up by the alter- 
nating current. 

Suppose now that we bring out two 
taps, so as to include 25 turns. We 
get 25 volts across the taps. Suppose 
we tap on so as to include 250 turns. 
We get 250 volts. We can connect 
out series motor to these taps (see 
Fig. 30). These rapidly reversing 
250 volts of self-induction will drive 
an alternating current of the same fre- 
quency through our series motor. This 
current, since it is driven by the volt- 
age of self-induction, will be in oppo- 
sition to the current from the alterna- 
tor. At a particular instant, for ex- 


\0>-^— iV 

FIG. 27. 

and in doing so cut across the loops 
again, but in the opposite direction. 
The greater the total magnetism and 
the greater the number of reversals 
per second, why the greater the volt- 
age induced in each loop. In other 
words, the greater the rate of cutting 
done by each loop, why the greater 
the voltage induced in the loop. We 
can assume in this case that the num- 

FIG. 2(). 

ample, the line current will be flow- 
ing down through 750 turns, will be 
opposed by the motor current flowing 
up through the 250 turns. The two 
currents will combine and flow out 
through the top tap to the motor, then 
down through the motor. The line 
current will go on into the line and 
back to the alternator. The current 
that came from the 250 turns will flow 
back through the bottom tap into the 
250 turns again. 



April, 1908 

We have just seen that the motor 
current is in opposition to the line cur- 
rent, consequently, it will have a de- 
magnetizing effect on the iron ring. 
This diminution in the magnetism will 
cut down the voltage of self-induction. 
This will allow a bigger current to 
flow from the line. The increased 
line current, however, is not sufficient 
to return the magnetism to its former 
value, for if it were the voltage of 
self-induction would regain its former 
value and the bigger current could 
not continue to flow. If we increase 
the load on the motor we draw a 
heavier current from the 250 turns, 
and accordingly wipe out still more of 
the magnetism in the ring, which al- 
lows a still bigger line current to flow. 
So the line current will go on in- 
creasing as the load on the motor in- 
creases. There is a definite ratio be- 
tween the line current and the motor 
current. In the case where the motor 
is taking, say, 200 amperes at 250 
volts, the alternator delivers 50 am- 
peres at 1000 volts. The ratio is 1000 
to 250, or four to one. Now if our 
coil had 10,000 turns the ratio would 
be 40 to one. So when the motor 
takes 200 amperes at 250 volts, the 
alternator gives five amperes at 
10,000 volts. As we increase the 
voltage we must also increase the 
quality of the insulation. There is no 
special difficulty in getting a high volt- 
age, say 75,000 volts, but there is 
great difficulty in finding an insulator 
that will stand up indefinitely against 
such a pressure. However, lines em- 
ploying 60,000 volts are in use. Now 
the generator voltage need be only 
one-tenth of the line voltage, for 
transformers can run up the voltage as 
high as desired. The motor voltage 
need be only 250 volts, for transform- 
ers can cut the line voltage down as 

You will note here that we can, by 
tapping the transformers at succes- 
sive points, get the graduated voltage 
required for starting the motor, and 
accordingly eliminate the rheostatic 

Now we can wind two coils on the 
same ring, and, as we have just seen, 
can, by sending an alternating current 
through one coil, induce an alternating 
current in the other. If one of these 
coils has 1000 turns and the other 
60,000 turns and the proper amount of 
iron is in the circuit we can, by apply- 
ing 1000 volts to the 1000 turns, get 
60,000 volts from the 60,000 turns. 
Consequently, we can, with such a 
transformer, use a 1000-volt dynamo 
and get a 60,000-volt feeder pressure. 
Likewise, if we had another trans- 
former of the ratio of 60,000 to 
10,000, we could step down from a 
feeder pressure of 60,000 volts to a 
trolley pressure of 10,000 volts, and 

with the auto transformer could step 
down from a trolley pressure of 
10,000 volts to a motor pressure of 
300 volts, or 200 volts, or 100 volts, 
or 50 volts — in fact, we could get 
whatever pressure was required for 
running or starting the motor. As- 
suming, for simplicity, that there are 
no losses in the transmission and that 
the motor requires 90,000 watts, we 
see that the generator will deliver 90 
amperes at 1000 volts to the step-up 
transformer, that the step-up trans- 
former will deliver 1.5 amperes at 
60,000 volts to the transmission line, 
and that the line will deliver 1.5 am- 
peres at 60,000 volts to the step-down 
transformer, which will deliver nine 
amperes at 10,000 volts to the trolley, 
which will deliver nine amperes at 
10,000 volts to the auto transformer, 
and if the motor is built for 250 volts 
the output of the auto transformer 
will be 360 amperes at 250 volts. 
The alternating current, as stated, 

fig. 30. 

will have no appreciable electrolytic 
effect. The single-phase series motor 
then is a solution of the electrolysis 
problem, a solution of the starting-loss 
problem and a solution of the trolley- 
voltage limit problem. It would seem 
then that the single-phase locomotive 
is almost ideal, and that it will have 
an enormous field of usefulness. 
However, the evolution of a commer- 
cial single-phase motor has taxed the 
talent of the designing engineer. We 
have just seen that alternating current 
tends to induce counter-currents. The 
problem of the designing engineer has 
been to keep these induction effects 
within the danger limits. If alterna- 
ting current were applied to the or- 
dinary series motor, why the induced 
currents would burn out the motor in 
short order. The motor must be so 
designed as to minimize the voltage of 
self-induction, and minimize the cur- 
rents due to the induced voltages. 
These induced currents generate heat. 
These heat losses are in addition to 
the losses incident to the direct cur- 
rent. The total losses of the alterna- 
ting-current motor, then, are greater 
than those of the direct-current mo- 
tor, and if the heat due to the total 
losses cannot be radiated as fast as 
they are generated, why the motor 
will eventually burn up. However, 
the skill of the designing engineer has 
finally triumphed, and we now have 

an alternating-current series motor 
that can compete with the direct-cur- 
rent motor. 

The single-phase locomotive, then, 
is a special case of the direct-current 
locomotive. The motors are series 
wound, but they are so designed that 
they minimize the induction effects of 
the alternating current. The start- 
ing rheostat of the direct-current mo- 
tor is replaced by the auto transform- 
er. The controlling device for opera- 
ting the switches is essentially the 
same as that used with direct current. 
The trolley arrangement is the same 
except that it is insulated for the high- 
er trolley voltage. The brake rigging 
is the same. The single-phase loco- 
motive, then, when approached from 
the standpoint of the steam locomo- 
tive and of the direct-current locomo- 
tive, is a simple affair after all. 


There is just one more point at 
which you should glance before we 
finish our remarks on alternating cur- 
rent ; it is the effect this self-induction 
has on the time relation of the main 
voltage to the main current. You 
know that our hottest days come after 
our longest days. Well, in a circuit 
in which there is self-induction the 
heaviest current drags behind the 
heaviest voltage and the amount of the 
lag varies with the nature of the load, 
the nature of the circuit, etc. How- 
ever, you do not need to worry about 
this, but if you will bear in mind that 
when a current is a rapidly reversing 
one, it will produce self-induction in 
any coil in the circuit, and the effect 
of the self-induction will be to make 
the time of the maximum current lag 
behind the time of the maximum volt- 
age, and you will see that the power 
at any instant is the product of the 
instantaneous current by the instan- 
taneous volts. So, if, when the volt- 
age is maximum, the current is only 
80 per cent, maximum, the actual en- 
ergy of the circuit is only 80 per cent, 
of the product of the maximum cur- 
rent by the maximum volts. The pres- 
ence of self-induction then compli- 
cates the measurement of alternating- 
current energy. But that does not 
concern us here. Knowing that self- 
induction and lag exist does not in- 
volve the necessity of your measuring 

It is essential, however, to measure 
the power the alternating-current lo- 
comotive takes to do its work. The 
watt meter will measure it for us. 
The watt meter is a diminutive motor 
whose speed is proportional to the 
power in the main circuit. It is so 
designed that its field strength is pro- 
portional to the line voltage, and its 
armature current to the line current. 
It will accordingly respond to varia- 

April, 1908 



tions in either, and measure the in- 
stantaneous product of both. The 
little motor is the accountant that 
charges up the locomotive with its 
drafts on the power-house. It repre- 
sents the commercial aspect of the 
case. It is the tell-tale that ultimately 
decides the fate of the machine. It 
is the final check that is applied to the 
industrial application of Faraday's 
discovery. Having arrived at the me- 
ter, we have reached the point where 
we can measure the power consump- 
tion of the single-phase locomotive, 
and that is as far as we need to go. 


I have given you no details as to 
the electric locomotive, nor is it my 
intention to do so. You will find such 
technical descriptions in the trade 
journals, and there will be plenty of 
specific articles dealing with your par- 
ticular locomotive. What you want is 
a grasp of the general principle that 
pervades the machine. You know that 
knowledge is increasing discrimina- 
tion. Consequently, if you understand 
the functions of the locomotive and 
add to that a mental picture of cur- 
rent-producing motion, and a mental 
picture of motion-producing current, 
why it is only a question of easy dis- 
criminating steps to arrive at the 
specialized case of the single-phase 
locomotive. As I have pointed out, it 
is mainly a question of patience. A 
single principle will explain a multi- 
tude of facts. The economical method, 
therefore, is to capture the principle 
instead of laboriously memorizing the 
facts. Now I have aimed to blaze the 
trail and have tried to make it a 
straight one. It is not a short one, but 
like any other path, the more often it 
is used the easier it is to follow. I do 
not know of any short cut. Locomo- 
tive work is our subject and the work 
must be measured. The magnetic 
properties of the electric current are 
the basis of the motor, consequently, 
induction must be understood. The 
commutator is a practical means of 
securing continuous armature rotation 
in a magnetic field, so it must be clear- 
ly understood. The rotation of the 
armature in the magnetic field sets 
up a voltage in the armature, conse- 
quently we must discriminate between 
the conditions incident to operation 
as a motor and operation as a dynamo, 
and see that the same current which 
makes the motor armature push also 
makes the dynamo armature drag. 
We also get a quantitative notion of 
the relation of voltage to speed, and 
combine this with the quantitative re- 
lation of current to torque in order to 
be able to understand that the product 
of current by volts means power. 
Knowing that current produces heat 
we must find a means not only of 

measuring the heat, but of calculating 
the current, the resistance through 
which the current flows, and the volt- 
age required ' to drive the current 
through the resistance. Having found 
out how to translate electric energy 
into terms of heat we must trace the 
chain of transformation from the coal 
to the motor in order that we may 
compare the performance of the elec- 
tric locomotive with the steam. Since 
the series motor gives the electric lo- 
comotive the same speed and draw- 
bar characteristics as the steam loco- 
motive, we must understand the series 
motor. Having compared the steam 
locomotive and the direct-current lo- 
comotive and discovered the limita- 
tions of direct current and the possi- 
bilities of alternating current, it is 
necessary for us to know how alter- 
nating current can be produced, trans- 
mitted, applied and measured, in order 
that we may know just where the 
single-phase locomotive stands with 
reference to the steam locomotive. 
Having arrived at that point, we have 
achieved the object of our talk — to 
put the electric locomotive before you 
as an understandable commercial ma- 

Long Acre Company Hearing 

The Public Service Commission, in 
its investigation of the electric light 
companies of New York, held a meet- 
ing March 12th to look further into 
the affairs of the Long Acre Co. 
Walter H. Knight stated that, of the 
total issue of $1,000,000.00 worth of 4 
per cent, bonds, $500,000.00 had been 
sold and $100,000.00 hypothecated as 
collateral with The American & Brit- 
ish Mfg. Co. Mr. Knight stated that 
this company had been given a formal 
order on July 3, 1907, to perform all 
work and labor in connection with 
the proposed Long Acre plant at the 
actual net cost, as shown by vouchers 
presented to and duly approved by the 
general engineer of the Long Acre 
Company, plus a profit of 15 per cent., 
the Long Acre Co. agreeing to sur- 
render $100,000.00 of the face value of 
its bonds to the American & British 
Co. for the performance of its terms 
of the agreement. 

During the year 1907 the Long 
Acre Co. had no receipts, and its ex- 
penditures were mostly for incorpora- 
tion matters. 

Henry W. King, as counsel for the 
Long Acre Electric Light & Power 
Co., stated in response to an inquiry 
respecting the books of the American 
Electric Illuminating Co. that he had 
traced the books of said company into 
the hands of John M. Ward, its re- 
ceiver, but that Mr. Ward was dead 
and he was unable to trace the books 
further. He failed to find the books 

of the American Electric Manufactur- 
ing Co. The question being raised as 
to whether the American Electric Il- 
luminating Co. had operated its fran- 
chises prior to July 1, 1907, Mr. King 
stated that this company operated a 
plant in a building on East 25th Street 
at Avenue A, with 250 arc-light dy- 
namos, stringing wires down Avenue 
A to Houston Street in 1889. 

The Long Acre Co. franchise was 
granted to the American Electric 
Mfg. Co. in 1887, was assigned in 
1888 to a Mr. Townsend and in 1889 
was assigned by him to the Ameri- 
can Electric Illuminating Co. In con- 
sequence of a judgment obtained 
against that company by a Mr. Dalton 
in 1897, the property was sold at auc- 
tion to a Mr. Minturn for $100.00, the 
only property being found to be the 
company's franchise. In 1896 Mr. 
Minturn turned the franchise over to 
the Long Acre Electric Light & Pow- 
er Co. The franchise is perpetual and 
applies to the City of New York as it 
was organized in 1887, which in- 
cluded the present Borough of Man- 
hattan and the portion of the Borough 
of The Bronx west of the Bronx 
River. One of the stipulations of 
this franchise was that one arc lamp 
was to be supplied to the city free for 
every 50 arc lights furnished to other 
customers, and Mr. King contended 
that the American Electric Illumina- 
ting Co. had fulfilled this stipulation, 
but admitted that it had not paid a 
cent per lineal foot for the streets 
which it occupied under the city per- 

An attempt on the part of Com- 
missioner Maltbie to elicit informa- 
tion regarding the real ownership of 
the company and the 490 shares of 
stock voted by Mr. Bouchie, only 
elicited information that Mr. Bouchie 
was the trustee for the Manhattan 
Transit Co., and that the other 10 
shares of stock were held by individ- 
uals for the purpose of qualifying 
them as directors. 

Mr. Knight, being recalled to the 
stand, stated that the Long Acre Co. 
had been engaged since January, 
1908, in constructing a small central 
power plant and placing ducts lead- 
ing to the conduits in the streets, and 
that it is now engaged in making sub- 
sidiary connections to its customers. 
The power plant is located on Second 
Avenue and Forty-seventh Street. 
The only customer connected at the 
time of the hearing was the Manhat- 
tan Transit Co., whose bills run from 
$400.00 to $500.00 a month. 

Mr. Knight stated that the company 
was planning a storage battery station 
and was carrying out plans for a large 
Waterside station, and that they had 
an offering of the lamp load of the 
Criterion Theater when its present 



April, 1908 

contract expired. The relationship 
between the Manhattan Transit Co. 
and the Long Acre Co. is that the 
Manhattan Transit Co. owns 98 per 
cent, of the capital stock of the Long 
Acre Co. and that the plant of the 
Long Acre Co. is located on the prop- 
erty of the Manhattan Transit Co., 
from which it leases room. 

Mr. W. H. King, on being recalled, 
gave the following inventory of the 
property of the Long Acre Light & 
Power Co., as of February 1, 1908: 

2 75-h.p. Diesel oil engines. 

1 5-panel main switchboard. 

2 6-in. k\v. generators. 
6 Wattmeters. 

1 10,000-gal. oil tank. 

2 10-h.p. motors and regulators. 
1 5-h.p. motor and regulator. 

I J /j-h.\). motor and regulator. 

1 lot of encased fuses. 

1200 ft. No. o lead-covered cable. 

Wiring and connections to cable. 

Cash in treasury. $18,010.24. mak- 
ing a total of $34,010.24. 

On being questioned as to the ad- 
vantage arising from competition to 
be introduced by his company, Mr. 
King stated that the consumer nat- 
urally profits by competition; that his 
company could sell electricity at a 
less price than the commonly prevail- 
ing price in New York. He did not 
consider it at all likely that there 
would be an agreement between the 
two companies to fix prices. Mr. 
King admitted that it was possible for 
the Consolidated Gas Co. to acquire 
a controlling ownership which would 
eliminate competition. 

Legal Notes; 


11 I i,l I WAV. 

A telegraph company obtained from 
a bridge company the right to lay its 
cables and wires across the bridge for 
an annual rental. Thereafter the 
bridge was acquired by the county in 
which it was situated. Held, that the 
county could not compel, by a suit in 
equity, the telegraph company to re- 
move all wires and cables or pay rental 
for the use of the bridge ; it having 
adequate remedy at law in an action 
to recover damages for the use while 
no compensation was paid. Beaver 
County v. Central Dist. & Print. Tel. 
Co. Supreme Court of Pennsvlvania. 
63 Atlantic 846. 


Under Rev. St. 1895. art - 3 OI 7- subd. 
2, giving a right of action for wrong- 

ful death, a complaint in an action for 
death against a corporation furnishing 
electric power for domestic, etc., pur- 
poses, which alleged that while dece- 
dent, an employee of one of defend- 
ant's customers, was repairing an elec- 
tric wire, defendant's engineer negli- 
gently and contrary to his agreement 
with decedent turned on the current, 
thus causing decedent's death, but 
which did not allege that defendant 
failed to exercise due care in selecting 
a competent engineer, or that it was 
negligent in the selection of its ma- 
chinery or in the construction of its 
poles and wires, did not state a cause 
of action. Williams v. Northern 
Texas Traction Co. Court of Civil 
Appeals of Texas. 107 Southwestern 


Where the wires connecting with a 
city lighting plant ran along a parti- 
tion wall above the roof of a building, 
it was the city's duty to the owners 
of the buildings and their servants, 
and to others having a legal right to 
use the roofs, to maintain such wires 
in a safe condition. City of Greenville 
v. Pitts. Supreme Court of Texas. 
107 Southwestern 50. 


Where an electric railroad company 
is authorized by a city to locate trolley 
poles along a street; they are not 
nuisances; hut such permission does 
not authorize the company to locate 
its poles so a> to unduly and unneces- 
sarily interfere with the public use of 
the street'-, or with the use of proper 
ways of ingress and egress to and 
from the street by abutting owners. 
Lambert v. Westchester Electric Ry. 
Co. Court of Appeals of New York. 
83 Northeastern ^)~~. 


Without regard to the respective 
contentions of the parties a- to the 

proximate cause of the electric light 
wire being down across the street 
which the plaintiff was traversing at 
the time that he came in contact with 
it. and was injured by the electric cur- 
rent with which it was charged, the 
jury could have found from the evi- 
dence that the proximate cause of the 
plaintiff's injuries was negligence on 
the part of the defendant in turning 
on such electric current, after receiv- 
ing actual notice that the wire was 
down, and without taking an}- steps to 
remove the dangerous situation which 
existed by reason of its being down. 
Mayor of Madison v. Thomas. Su- 
preme Court of ( ieorgia. 60 South- 
eastern 461. 


Though the power to establish a 
tariff of rates for a public service cor- 
poration is legislative, and not an ex- 
ecutive or judicial function, the Legis- 
lature may delegate the right to fix 
rates for a specified service toan admin- 
istrative body to conform to a stand- 
ard established by the Legislature, 
especially where it appeared that no 
uniform rate of charges could be es- 
tablished throughout the State that 
would be just or reasonable, and that 
an approximation of a reasonable 
tariff would require special rates to 
be prescribed for many different lo- 
calities. / r illage of Saratoga Springs 
v. Saratoga Springs Cas, Elect. Lt. & 
P. Co. Court of Appeals of New York. 
83 Northeastern 693. 


Laws 1905. p. 2092, c. 7 2,7, estab- 
lishing a commission of gas and elec- 
tricity, and authorizing it to fix. after 
hearing, within the limits prescribed 
by law, the maximum price for gas 
and electricity furnished by any public- 
service corporation, is not a violation 
of the federal Constitution guarantee- 
ing to every State a republican form 
of government, in that it assumes t<> 
delegate legislative powers to an ad- 
ministrative body ; the true meaning 
of the constitutional division of gov- 
ernmental power being that the 
"whole" power of one of the three de- 
partments of government shall not be 
exercised by the same hands which 
pnssc^ the '"whole" power of either 
of the other departments, there being 
no objection to the imposition on an 
administrative body of some powers 
legislative in character. / 'illagc of 
Saratoga Springs v. Saratoga Cas. 
El. Lt. & P. Co. Court of Appeals "i" 
New Y\ork. 83 Northeastern 693. 


Where plaintiff, a police officer, 
went on the roof of a building in the 
nighttime to detect persons violating 
the law. without the owner's knowl- 
edge or consent, he was at most a 
licensee, and could not recover against 
the city for injuries received from con- 
tact with an improperly insulated elec- 
tric wire belonging to the city, since 
it was under no duty to plaintiff to 
keep the wires in a safe condition. 
City of Greenville v. Pitts. Supreme 
Court of Texas. 107 Southwestern 

April, 1908 



A New Type of Induction Motor 

The flexibility and economy of the 
alternating current method of power 
distribution has led to its almost uni- 
versal adoption in central station 
service, and because of the economy 
of the use of this central service many 
industrial plants are now purchasing 
power. The same features that led to 
the selection of alternating current 
generation and distribution for central 
stations have, in a somewhat lesser 
degree, led to its distribution for large 
industrial works. On account of this 
it was necessary to provide alternat- 
ing-current motors of varying char- 
acteristics. The earlier polyphase 
motor was of the brushless or squirrel 
cage type and compared more nearly 
with the shunt-wound direct-current 
motor. To meet the varying condi- 
tions imposed by high starting torque 
with a low initial starting current, a 
modified form of motor has been 
produced by the Westinghouse Elec- 

tinuous rings of metal so that there is 
no occasion whatever for sparking. 
The brushes are liberally proportioned, 
so that they have long life and re- 
quire the minimum of attention. The 
rotor is unlike the usual short-cir- 
cuited type of armature on induction 
motors, and has a regular winding in 
which the resistance is inserted while 
starting, and thus the starting current 
is limited. Approximately full load 
current is required to produce full 
load starting torque. 

Special covers are provided for en- 
closing the brushes where this may 
be desired. The illustrations show the 
method employed with the semi-en- 
closing covers by which the brushes 
and slip rings are protected from any 
coarse particles, and the totally enclos- 
ing covers which prevent all dust 
from lodging on these parts. 

Type HF motors are built in the 
usual sizes from 5 to 500 h.p., with 



trie & Mfg. Company known as the 
type HF. 

The general appearance of this 
motor is shown in the illustration, 
from which it will be noted that it is 
especially rugged and substantial. 
The frames are amply ventilated, but 
so constructed as to protect the lam- 
inations from injury. Throughout, 
the motors have been designed for 
hard service. The insulation on the 
coils has been very thoroughly made 
and will stand much greater stresses 
than it receives in normal service. It 
is furthermore so built that vibration, 
and consequent wear of the insulation, 
is prevented. 

Attention is called to the brushes 
shown in the illustration, by means of 
which connection is made to an ex- 
ternal resistance. These bear on con- 

larger sizes on special order. TTie 
standard frequencies are 25, 40 and 
60 cycles, but motors for other fre- 
quencies are supplied when required. 
In connection with these motors, 
special starting devices are supplied 
by which the current at starting i- 
varied by hand. These starters cut 
resistance out of the circuit of the 
rotating armature until the handle 
reaches the full-on position, when the 
resistance is all short-circuited and the 
motor runs with practically the con- 
stant speed characteristics of the 
squirrel cage armatures. A special 
short-circuiting switch is provided on 
motors of 100-h.p. or larger if de- 
sired. This relieves the brushes and 
.the leads and contacts on the starter. 
as the windings are short-circuited in- 
side the motor. 

Aids to the Solution of Practical 
Illuminating Problems 

The science of illuminating engi- 
neering is as yet in its infancy, and 
data of value in the laying out of 
lighting installations are not plentiful. 
I he activity of the General Electric 
Company in the advancement of the 
art of illumination is evidenced by its 
systematic dissemination of informa- 
tion on this subject. Two bulletins 
recently compiled by the Harrison. 
X. J.. work> of the company illustrate 
the thoroughness with which such 
subjects are now being treated. These 
bulletins are valuable primarily to the 
practical man engaged in planning 
lighting installations and are also of 
interest to every central station or- 
ganization as a whole. 

A perusal of these bulletins, Nos. 
4561 and 4506 on GEM high efficiency 
incandescent units and tantalum in- 
candescent lamps, respectively, will 
convince the reader of their value as 
aids to the solution of practical illum- 
inating problems. The illumination 
tables contained in these bulletins are 
invaluable in such work and contain 
information hitherto not available. 
This data leads itself readily to the 
laying out of lighting installations and 
will be appreciated by architects, con- 
tractors, solicitors, and others daily 
confronted with the solving of illum- 
inating problems. 

In bulletin No. 4566 the tantalum 
lamp with its high efficiency and long 
life is specially recommended to 
central stations as a desirable factor 
in the reducing of peak load condi- 
tions inasmuch as the use of the 40 or 
80-watt tantalum lamps gives the cus- 
tomer 2^ per cent, more light than the 
standard 16 or 32-c-p. lamps, and with 
an expenditure of 20 ner cent, less 
energy on the part of the central 
station. Thus it is .seen that the gen- 
eral satisfaction which attends the use 
of the tantalum lamp is shared alike 
by both buyer and seller of current. 
Stress is laid on the brilliant and at- 
tractive quality of light emitted by the 
tantalum lamps, qualities which make 
it mot desirable for hotels, theaters. 
cafes, stores and all public building's. 

The use of the table showing the 
actual amount saved in dollars and 
cents by the use of these lamps should 
prove of great value to lighting solicit- 
ors as a convincing argument in favor 
of die more brilliant and efficient tan- 
talum lamps. This table shows that 
the tantalum lamp at the average 
lighting rates now in force will save 
more than twice the initial cost (hiring 
its average life of 750 hours. 

Several pages are devoted to the 
elucidation of a practical method for 
the solution of illuminating problems. 
The illumination table intended for 



April, 1908 

use in connection with this method is 
comprehensive for the 8o-watt lamps 
with three types of holophane reflect- 
ors, and by interpolation can also be 
used for the 40-watt and 50-watt 
tantalum units. Several curves giving 
the candle-power distribution with dif- 
ferent types of holophane reflectors 
are also shown. 

Bulletin Xo. 4561 contains valuable 
information 111 regard to "GEM" high 
efficiency incandescent units. These 
lighting units are composed of the 
"GEM" lamps with varying types of 
holophane reflectors and are made in 
100, 125, 187 and 250-watt sizes. 
Candle-power values of these lamps 
are not given, as it is obvious that 
these values will vary according to the 
type of reflector used. 

The method of solving illuminating 
problems which was referred to in 
l3ulletin No. 4566 is here treated more 
fully. The subject of illumination in 
general is taken up and treated briefly. 
This is followed by a specific treat- 
ment of the subject with reference to 
"GEM" high efficiency units. Assum- 
ing the degree of illumination desired, 
a problem is worked out, reference 
being made to the illumination tables 
for the solution. 

Two photographs are shown of the 
interior of a dry-goods store before 
and after the installation of Gem 
units. The original installation con- 
sisted of six-light electroliers with 
ordinary 10-c-p. incandescent lamps, 
placed yYx feet from the floor. The 
Gem installation consisted of 125-watt 
high candle-power units with bowl 
holophane reflectors. These units were 
installed on the ceiling 14 feet above 
the floor. 

The Bristol Company 

The Bristol Company, of Water- 
bury, Conn., has come under the con- 
trol of Prof. William H. Bristol whose 
inventions this company has been 
manufacturing since it was first or- 
ganized in 1889. Prof. Bristol as- 
sumed active charge of the manage- 
ment of the business on Friday. March 
28th, and now owns the majority in- 

The business which has been car- 
ried on under the personal name of 
Wm. H. Bristol at Xew York will 
hereafter be combined with The Bris- 
tol Company, and by this consolida- 
tion of interests The Bristol Com- 
pany will now have the most com- 
plete line of recording instruments in 
the world for pressure, temperature, 
electricity and for a great variety of 
other applications. 

The Bristol Company was organ- 
ized in 1889 under the name of 
"Bristol's Manufacturing Company" 
to manufacture Bristol's pressure 

gauges and Bristol's steel belt lacing, 
for which Wm. H. Bristol had taken 
out patents. To these were added 
many other inventions from time to 
time, and in 1894 the business was 
incorporated under the name of "The 
Bristol Company." 

Two years ago Wm. H. Bristol 
withdrew from the presidency of the 
company, and since that time has de- 
veloped many new inventions, includ- 
ing the Wm. H. Bristol electric pyrom- 
eters and patented smoke chart re- 
corders. The new pyrometers have 
come into wide use, there being, for 
instance, fifty of these pyrometers in 
service in one of the large steel plants. 

Mr. Bristol has taken out a large 
number of patents during the last 
three years on new instruments. One 
of these which will be soon put on the 
market is the long distance electric 
thermometer, designed especially for 
indicating and recording refrigeration, 
atmospheric and drying temperatures. 
This instrument will fill a long-felt 
want for use where it is desired to 
quickly indicate at some central sta- 
tion by means of switches the tem- 
peratures at several distant points. 

The new lines of Wm. H. Bristol 
instruments supplement those of The 
Bristol Company, supplying a variety 
for applications for which the old in- 
struments could not be recommended. 
For example, the standard Bristol's 
recording thermometers cannot be 
successfully used for temperatures 
above 6oo° F.. while the Wm. Ff. 
Bristol pyrometers are being applied 
to great advantage for the higher 
ranges of temperature, especially for 
ranges from 6oo° to 2600 F. 

The new lines of Wm. H. Bristol 
pyrometers are fitted with special 
movements made by The Weston 
Electrical Instrument Company, and 
are designed for extremely accurate 
measurements. The combined line of 
recording instruments to be hereafter 
manufactured by The Bristol Com- 
pany will make it possible for the 
company to co-operate better than 
ever before with its customers in giv- 
ing them perfectly satisfactory service. 

importance ; i. e., for marine use and 
also for power-houses and establish- 
ments where the economy of floor 
space is an important consideration. 
The pump cylinders are of the piston 

A. New "Vertical Pump 

For many years there has been a 
growing demand for an inexpensive 
vertical steam pump of small and mod- 
erate size adaptable to boiler feeding 
and general service, and where the 
modern high working pressures ne- 
cessitate substantial construction. The 
Blake & Knowles Steam Pump Works 
of Xew York City have just per- 
fected a new design for a vertical 
duplex pump which meets these con- 
ditions and which is nicely shown by 
the accompanying illustration. It is 
especially adapted for services where 
compactness and strength are of prime 


pattern and fitted with substantial 
brass linings. The pump pistons are 
very deep and packed with fibrous 
packing suited for hot or cold water. 
The piston rods are of Tobin bronze. 
The steam cylinders are of the regular 
duplex pattern and of similar design 
to that which has for many years been 
successfully used with the well-known 
horizontal Blake-Knowles special du- 
plex pumps. The cast iron cradle or 
center piece which ties the steam and 
water ends is- extremely rigid, an im- 
provement on the ordinary tie bar con- 
struction, as it prevents any possibility 
of the cylinders getting out of align- 
ment. The cylinders are fitted with 
brackets for bolting to a bulkhead or 
wall, although if preferred a special 
base is fitted for placing the pump 
directly on the engine room floor. The 
pumps are suited for a working water 
pressure of 200 lb. per sq. in. A new 
illustrated publication, B-K 811, is- 
sued by the manufacturers, contains 
complete information with table of 
sizes and dimensions of this type of 

April, 1908 



Steam Turbine Sales 

We publish herewith tables giving 
a summary of steam turbine sales of 
the Curtis vertical turbine and the 
Westinghouse horizontal turbine. The 
figures are up to Dec. 31, 1907, from 
the beginning of the steam turbine era. 

A number of interesting facts are 
revealed in the statement of Curtis 
turbine sales. 

Curtis turbine generators for the past 
fiscal year of the General Electric 
Company, 286,320 kw. capacity, or 
more than 25 per cent, of the total 
sales since the Curtis turbine was in- 

Another fact of considerable inter- 
est is the large number of plants for 
which the Curtis turbine has been 
selected as prime mover. The large 


to Dec. 31, 1907 

Number of Plants 

1000 kw. 

and less 

Central Station and Railway Traction 
Industrial Plants and Miscellaneous.., 





1000 kw. 








kw. cap. 












kw. capacity 
per machine 




Installations to Dec. 31 , 1907 




Orders on hand Dec. 31, 1907 


Total Sales to Dec. 31, 1907 



1,073 695 






Plant Capacity 

up to 






Electric Traction 

Electric Railways 

(R.R. Electrification) 


Electric Lighting 

Central Stations 

Isolated Plants 



Steam Railroads 

R. R. Electrification 

(R.R. Car Shops) 



Textile Mills 

R.R. Car Shops 

Cement Mills 

Iron and Steel Works. . . 
Pulp and Paper Mills. . . . 

Rubber Works 

Powder Works 

Machinery Manufacturers 
General Manufacturers. . 


Mining and Irrigation 

U. S. Government 


Grand Total 














5 400 







' 4,766 





Av. ca- 
pacity of 














10 500 


























192 000 



1,012— 7 
621— 7 
400— 3 











Industries in parentheses, but allowed for in grand total. 

Note. — Business uncompleted December 31, 1907: 60 turbines of 153,550 kw., total, leaving shipped or 
in operation 433 machines, or 487,150 kw., averaging 1,122 kw. 

The most noticeable single item is 
the total capacity sold to Dec. 31, 
1907—1,073,695 kw., or about 1,556,- 
000 brake h.p. This is the strongest 
indication of the advance of the 
steam turbine generating unit that 
has ever been published. That this 
advance is accelerating rapidly is 
shown by the amount of the sales of 

range of sizes in which this turbine 
is sold is probably responsible for the 
great variation in average sizes of 
plants in which it is used. The large 
central stations and electric traction 
enterprises with an average size of 
3778 kw. plant capacity strikingly 
differ from the industrial plant of 305 
kw. average capacity. 

New A.llis - Chalmers Alternator 
for Nevada - California Power 
Company, Goldfield, Nevada 

The Nevada- California Power 
Company, formerly the Nevada 
Power, Mining and Milling Company, 
Goldfield, Nev., is preparing to install 
a fourth Allis-Chalmers alternating- 
current generator,of the water-wheel 
type, having 1 500 kw. rated capacity, 
to augment the enlarged power service 
now contracted for in the vicinity, 
pending the completion of their new 
hydro-electric stations on Bishop 
Creek, Inyo County, California. 

The unit is a 3-phase, 60-cycle, 
2200-volt machine to run at 400 rev. 
per min. and is arranged for direct 
connection to a water-wheel operating 
under an 850 ft. head, alongside of 
three similar units already installed 
and similarly driven, the last of which 
was placed in service in 1905. 

These generators are of the stand- 
ard Allis-Chalmers water-wheel type, 
with two bearings and extended shaft. 
Current is transmitted 113 miles at 
60,000 volts to Tonopah and Goldfield, 
Nev., with branches from this line to 
other points, and supplies power to a 
number of the mines in that district 
whose equipment is electrically oper- 
ated. The following are a few of the 
properties in Tonopah and Goldfield 
which have installed Allis-Chalmers 
induction motors ranging in size from 
5 to 100 h.p., and are supplied from 
the Nevada-California Power Com- 
pany's mains : Goldfield Consolidated 
Mines Co., Goldfield Milling & Mfg. 
Co., Nevada Goldfield & Reduction 
Co. and Montana Tonopah Mining Co. 
The latter company has installed fif- 
teen motors aggregating 600 h.p. 
With the completion of its two sta- 
tions, the Nevada-California Power 
Company will possess hydro-electric 
power facilities unequaledin this sec- 
tion, as there are two plants already 
in operation, and the combined ca- 
pacity will be 14,000 h.p. Much ad- 
ditional power can also be developed, 
as the company controls 3,200 ft. of 
fall on Bishop Creek, and large 
storage reservoirs can be built on the 
headwaters of the stream at no great 

Incandescent Lamps for Singer 
Building', New "YorK 

It has been announced that the new 
Singer Building, New York, will re- 
quire some 25,000 lamps to fill the 
sockets, and the first shipment of 10,- 
000 lamps has been made. Two hun- 
dred and thirty-volt Columbia lamps 
will be used, the contract having been 
secured by the Central Electric Com- 
'pany of Chicago, general Western 
sales agents for the Okonite Company. 



April, 1908 

Trade News Items 

The Northern Engineering Works, 
of Detroit, have furnished the Mam- 
moth Copper Mining Co., of Kennett, 
Cal., with a 15 ton, 48 foot span North- 
ern traveling crane. 

The Northern Engineering Works, 
Detroit, Mich., report recently ship- 
ping to the Nederlandsche Gist en 
Spiritusfabrik, Delft, Holland, an 
overhead track system, consisting of 
approximately 500 ft. of overhead 
track with hangers, switches, etc., and 
an electric one-ton trolley hoist, two- 
motor alternating-current design. 

The Chas. J. Bogue Electric Co., 
will remove on May 1st from 213 
Centre Street to 513-515 West 29th 
St.. New York. 

The telegraph, telephone and elec- 
tric-light companies reported the pur- 
chase of 3,493,025 round poles exceed- 
ing 20 ft. in length in 1906. < >ver 
three-fifths of these poles consisted of 
cedar and more than 28 per cent, of 
chestnut. Relatively small amounts 
of pine, cypress, redwood and other 
poles were also purchased. In addi- 
tion to the poles required by these 
commercial companies, a large num- 
ber of smaller poles were used for 
local telephone lines and similar pur- 

The Electric Cable Company, of 17 
Battery Place, New York, whose plant 
at Bridgeport, Conn., was partially 
destroyed by fire, announce that they 
have made arrangements which will 
permit of filling all orders. Pending 
adjustment of insurance details, the 
Company will make no announcement 
of its plans for re-building. 

The semi-annual meeting of The 
American Society of Mechanical En- 
gineers will be held in Detroit, Mich., 
June 23-26. Among the papers to be 
presented at this session are A Method 
of Cleaning Gas Conduits, by W. D. 
Mount; A Method of Checking Coni- 
cal Pistons for Stress, by Prof. George 
H. Shepard ; Clutches, with special 
reference to automobile clutches, by 
II. Souther ; Horse-power, _ Friction 
Losses and Efficiencies of Gas and Oil 
Engines, by Prof. L. S. Marks; Some 
Pitot Tube Studies, by Prof. W. D. 
Gregory ; The Thermal Properties of 
Superheated Steam, by Prof. R. C. H. 
Heck; A Journal Friction Measuring 
Machine, by Henry Hess; A By-Prod- 
uct Coke Oven, by W. H. Blauvelt ; 
Tests of Some High Speed Steam En- 
gines, by F. W. Dean. 

The executive offices of the West- 
inghouse Electric & Manufacturing 

Company, now at in Broadway, New 
York, N. Y., and the New York sales 
offices and export offices, of that Com- 
pany, now at n Pine Street, have 
been removed to the new City Invest- 
ing Building, 165 Broadway, New 

The general offices of the General 
Electric Co., of the American Loco- 
motive Co., and the New York sales 
office of the Crocker Wheeler Co., 
have been removed to the Hudson 
Terminal Building, 30 Church Street, 
.Yew York. 

The Wdieeler Condenser & Engi- 
neering Co., Carteret, N. J., has made 
arrangements with Charles S. Lewis 
& Co., Granite Building, Fourth and 
Market Streets, St. Louis, Mo., to 
handle "Wheeler" apparatus in the 
State of Missouri. 

The Westinghouse Machine Com- 
pany, builders of the Roney mechani- 
cal stoker, reports for the first two 
months of this year a good demand 
for stoker equipment. The Mer- 
chant's Ice & Cold Storage Company, 
of Cincinnati, are installing three 
equipments; the Montgomery lee & 
Cold Storage Company, Jenkintown, 
Pa., two equipments; and the Molt 
Ice & Cold Storage Company, Indian- 
apolis, Ind., the same number. An- 
other municipal lighting station at 
Troy, O., operated by the Hoard of 
Public Service, has adopted Roney 
stokers, and the Pennsylvania Light. 
Heat & Power Company, of Alle- 
gheny. Pa., operating a large central 
station at that place, increases its pres- 
ent equipment with two stokers of 
large capacity. The Mutual Union 
Brewing Company, Alliquippa, Pa., 
is installing three equipments, and the 
Alpha Portland Cement Company, 
Operating a number of plants in the 
country, one of the largest size. 

The Duquesne Steel Foundry, 
which operates a large plant in the 
Pittsburg district, has decided to 
adopt the gas-power system to operate 
the works formerly driven by steam. 
The initial equipment will consist of 
a 400 h.p. (max.) Westinghouse gas 
engine of the trTree-cylinder vertical 
enclosed type, direct connected to a 
240 kw. generator which will serve 
the various motor drives around the 

In the United States Circuit Court 
of Appeals for the Third Circuit, 
the Westinghouse patent Xo. 582481, 
issued to Nolan. assignor, for 
fastening the laminae of core plates, 
was upheld in a suit brought by the 
appellee against the Prudential Life 
Insurance Co. 

Catalogue Notes 

I iraded shunt resistance multigap 
lightning arresters for 1908 are de- 
scribed in a new bulletin issued by 
the General Electric Company, Sche- 
nectady, X. Y. The bulletin also 
contains detailed descriptions of low 
voltage arresters, static dischargers, 
constant-current horn arresters, dis- 
connecting switches, choke coils and 
the well-known Type M, Form D-2 
direct-current arrester for voltages up 
to 6000. Tables of general data re- 
garding the apparatus, connection and 
dimension diagrams, etc., are included. 
'Ldie multigap arresters for high volt- 
ages consist essentially of a series of 
knurled cylinders placed closely to- 
gether, the discharge taking place 
across the path of gaps thereby pro- 
duced and being extinguished before 
the dynamic current can follow it for 
more than half a cycle by reason of the 
peculiar composition of the metal 
making up the cylinders. 

The arcs are shunted by low, me- 
dium and high resistances which have 
the effect of making the arresters sen- 
sitive to a very wide range of fre- 
quencies and are claimed to discharge 
with safety under practically all con- 

Tlie multiple connection of the ar- 
resters allows them to relieve strains 
between line and line as well as 
between line and ground. An impor- 
tant feature of their design is the 
absence of series resistance which 
gives a free discharge at high fre- 
quencies as well as a free discharge 
of large quantities of lightning. The 
bulletin discusses briefly the theory 
upon which the arresters operate and 
their details of construction. It is 
b< mnd iu an artistic cover. 


In Bulletin Xo. 4561, the General 
Electric Company, Schenectady, 

X. Y.. has issued an interesting lC>- 
page pamphlet on illumination and the 
best solutions of many of the prob- 
lems involved. Illumination tables 
giving the foot-candle values on 
different horizontal planes when the 
lamp is used with various types of 
Elolophanes are also given. The 
"Gem" unit is easily renewed, fits the 
standard socket and burns on any 
standard circuit. It gives a perfect 
distribution of light with a downward 
efficiency of from \Y\ to nearly one 
watt per candle, and the quality 
of light is brilliant, soft and uniform, 
making it especially useful for all in- 
terior lighting. The "(Lin'* units are 
inexpensive, giving low investment 
charge and cheap renewals, and no 
repair account is necessary with their 


Volume XXXIX. Number 5. 
$1.00 a year; 15 cents a copy 

New York, June, 1 908 

The Electrical Age Co. 
New York and London 

Published monthly by The Electrical Age Co., 45 E. 42d Street, New York. 

J. H. SMITH. Pres. C. A. HOPE. Sec. and Treas. 


Telephone No. 6498 38th. Private branch exchange connecting all departments. 
Cable Address — Revolvable, New York. 


United Slates and Mexico, SI. 00; 

Canada. $1.50. To Other Countries, $2.50 


Insertion o( new adTertisemenls or changes of copy cannot be guaranteed for the following 
issue if received later than the 15th of each month. 


Business and Editorial Departments, 33 Bedford Street 

Strand. W. C. 

London Telephone, Post Office. 3807 Central. Private branch exchange. London Cable 

Address — Revolvable, London. 
All remittances to London Office should be crossed London and County Banking Company. 


Subscriptions may be sent to the following branch offices or agencies, where they will re- 
ceive the same careful attention as at the home office in New York : 

124, Queen St., Melbourne 33. Loop St.. Cape To«n 

359. George St. . Sydney Unter den Linden. 5. Berlin 

61, Main St , Yokohama Nevsky Prospect. St. Petersburg 

23, Esplanada Rd., Bombay 31. bis rue de Fauburg Monlmartre. Pari* 

General Agents for United States and Canada : The American News Company 

General Electric Company's 

The sixteenth annual report of the 
General Electric Co.. dated Jan. 31, 
1908. is remarkable as an exposition 
of a continued far-sighted policy of 
ultra conservatism in internal affairs. 
together with an equally notable policy 
of liberalism during troublous times 
in dealing with the vast aggregation 
of industries represented by 21.000 
regular customers. This was a policy 
intended to minimize the effect of hard 
times, and not to "kill the goose that 
lays the golden eggs." 

The evidence of this is found in the 
item of $3 1, 957,999-73 , shown as due 
from customers on Jan. 31. 1908. 

Of this sum $9,396,242.59 is in the 
form of notes, and $22,561,757.14 m 
open accounts. This sum remains clue 
after making cash collections of about 
S66. 000.000. although the sum re- 
maining due is only about $3,500,000 
more than was due the year before at 
the same date : it really represents a 
more important item. 

It is known that the business of the 
early part of the year was at a rate of 
885.000,000 a year. When the slump 
came, sales fell off to such an extent 
that the total of orders received for 
the year was but $59,301,040. and the 
total sales billed was $70,977,168. 
This, however, was an increase of 
18.2^ over the previous year. 

"Of some 21,060 regular custom- 
ers." the report says, "an increase of 
1300. there were debit balances against 
10.000 at the close of the year." 

It is not difficult to read between 
the lines and to discover a broad- 
minded policy carried out which kept 
the wheels of other industries going 
by not withholding the needed sup- 
plies and equipment, although deliv- 
eries involved the taking of extraor- 
dinary risks for collections. These 
risks have been provided for bv writ- 
ing doAvn the asset value of accounts 
receivable by the sum of $2,100,272.89. 

The conservative internal policy re- 
ferred to is evidenced by those por- 
tions of the report which show the 
writing off of millions in items which 
most companies carry along year after 
year as book assets. The first of 
these items is a sum of $872,345.67 
expended during the year in acquiring 
patents, licenses under patents and in 
patent litigations. This is in accord- 
ance with a policy which leaves all 
the company's patents, franchise-; and 
good-will standing on the books at 
a valuation of only $1.00. 

The same policy has been followed 
in dealing with investments in factory 
properties and materials on hand. 

The sum of S3. 745,989. 06 was 
written off during the year from the 
cost of factory properties, and the 
book value of the stock inventories 
advanced by $2,000,000, bringing the 
reported value of all material in hand 
below the market values on Jan. 31, 
1908, regardless of cost. 

The total sums written off the cost 
of factory properties during the last 
15 years has been $21,951,013. 93, 
leaving these properties valued at 
only $ 1 2.000.000. This represents a 
cost chargeable to manufacturing of 
only $2.00 a square foot of floor space. 

That this plays an important part 
in enabling the company to compete 
for the world's business is evident 
when it is considered that the average 
cost in the large electrical factories of 
the world is between $4 and $6 per 
square foot of floor sr5ace. 

Almost the entire cost of the Pitts- 
field plant of the Stanley Works, 
which was forced upon the company 
in 1903, has finally been liquidated. 

Notwithstanding the large sums 
written off. the declared profits of the 
company for the year were $6,586,- 
() ?.}-37- From this, dividends amount- 
ing to $5,183,014 were paid during 
the year, and an amount of $1,403,- 
039.37 was carried to the surplus ac- 
count, making a total surplus of $16.- 
5 ! 3-836. 14. Another item which at- 

tracts attention is the sum of $12,250.- 
720.92 cash in hand on Jan. 31, 1908. 

This large sum represents the cash 
which was provided in 1907 by the 
issue of $12,872,750.00. in 5 per cent, 
bonds, to provide ample working cap- 
ital for the great volume of business 
then in sight. As business fell off in 
the latter part of last year, this sum 
returned to the treasury and is now 
in hand, ready to be employed when- 
ever a revival of business may require 

Comment has been made upon the 
apparent falling off of the percentage 
of manufacturing profits for the year. 

The profits are said to have been 
actually reduced about 1 per cent, by 
the high price of materials and other 
temporary causes, but the other 
change is entirely accounted for by 
the conservative course of the man- 
agement in putting all of the assets 
on a basis which cannot apparently 
be adversely affected, even if the re- 
vival of business is delayed even be- 
yond the most pessimistic expecta- 

Reference is made in the report to 
the purchase of 700 acres of land 
near Erie. Pa., at a cost of $232,- 
301.53, to provide for future develop- 
ment at some point nearer the central 
West. "In view of the existing de- 
pression." the report says, "the erec- 
tion of buildings thereon is deferred 
for the present." 

Downward vs. Horizontal 


In the April issue oi The Elec- 
trical Age there is an article by Mr. 
T. H. Amrine, in which the question 

of the relative advantages ^<\ distribu- 
tion, as between the commercial forms 
of incandescent lamps, is discussed. 
The writer falls into an error which 
is undoubtedly common enough among 
laymen, but one which it is important 
that those professionally interested in 
illumination, at least, should avoid. 



June, J90& 

This error consists in considering the 
area of the ordinary distribution curve 
as significant of quantities of light. 
This error is as natural as it is wide 
of the truth. So natural is it to com- 
pare the areas of distribution curves 
that it often requires careful and 
elaborate explanation to convince the 
untechnical observer that the area has 
no significance whatever and that only 
the length of the radial lines must be 
taken into consideration. 

One of the stock arguments against 
the type of incandescent lamp, of 
which there are several commercial 
varieties, which aims to give a greater 
amount of light in the vertical direc- 
tion than the ordinary type, is the light 
lost in the base. Mr. Amrine attempts 
to show this by shading portions of 
polar diagrams of the so-called "down- 
ward light" type and the familiar oval- 
anchored filament type. A little con- 
sideration of the mathematical prob- 
lem involved will show that this well- 
worn argument has almost nothing to 
rest upon, and should have been worn 
out and discarded long ago. The fal- 
lacy of the argument may be summed 
up in two points : In the first place, 
the quantity of light intercepted by the 
base of the lamp is so small as to be 
practically negligible ; second, a con- 
siderable portion of this light is re- 
flected by the glass and cement with 
which the base is attached. 

A glance at the familiar Rousseau 
diagram will show how extremely 
small in quantity are the amounts of 
light included within given angles near 
the poles, as compared with similar 
angles near the equator. Thus, the 
amount of light falling within the 
15 degrees from the vertical is less 
than 2 per cent, of the total flux, as- 
suming the intensity to be equal in all 
directions ; and this amount un- 
doubtedly exceeds the total amount of 
light actually intercepted by the base 
of the lamp. The absorption-in-the- 
base argument may, therefore, be 
dropped as wholly academic and im- 

As to the relative merits of the dif- 
ferent types of lamps when used with 
reflectors, that is quite a different mat- 
ter, and, as in the case of the lamps 
used bare, will depend largely upon 
the particular type of distribution 
wanted. If a wide distribution is de- 
sired, unquestionably the lamp hav- 
ing a naturally wide distribution is 
preferable, since it is difficult to de- 
flect rays that are included, say. with- 
in the vertical and 45 degrees, into a 
direction nearer the horizontal. On 
the other hand, if vertical concentra- 
tion is desired, then the lamp having 
naturally a greater intensity in the 
vertical is easier to handle. 
There are. however, a considerable 

number of cases in which accessories 
by way of shades and reflectors are 
undesirable, as, for example, where the 
risk of mechanical breakage is great, 
or the presence of smoke or dust ren- 
ders them unfit for their purpose. In 
such cases lamps having a high verti- 
cal intensity will have the advantage 
for most purposes where incandescent 
lamps would naturally be used. 

The advocate of the standard type, 
that is, the type giving a horizontal 
intensity 'practically twice the vertical, 
will, of course, call attention to the 
fact that the vertical, or so-called 
"downward type," gives equally high 
intensity upward, barring the absorp- 
tion by the base. While this is true, 
the argument applies just as well to 
any other type of lamp. In other 
words, all types of lamps give prac- 
tically the same quantity of light above 
as below the horizontal plane passing 
through the filament, that is, divide 
their light equally between the upper 
and lower hemispheres. Since, in the 
case where lamps are used bare, the 
light in the upper hemisphere may 
generally be considered useless, the 
argument, as between the two, narrows 
itself down to the distribution in the 
lower hemisphere. 

In choosing between the several 
types of incandescent lamps, then, the 
points to be considered are : 

First — Is the use of a reflector pos- 
sible or advisable? 

Second — If a reflector is to be used, 
is a wide or concentrated distribution 

Third — If reflectors are not to be 
used, is a general distribution, or spe- 
cial lighting required? 

In any case, let it be borne in mind — 
First, that the particular form in which 
the filament is looped or placed in the 
lamp has absolutely nothing to do with 
the efficiency of the lamp. You can- 
not get something for nothing by 
simply giving a peculiar twist to the 

Second, a lamp that is giving 
greater intensity in some particular 
direction than some other form of 
lamp must, of necessity, be giving 
less intensity in some other direction. 
It is merelv a matter of deciding: in 
which direction it is most desirable to 
have the intensity. 

Third, the polar, or distribution 
curve, as commonly given, indicates 
how the light is distributed on the 
average in a vertical plane, and is to 
be read as a curve only, the area 
included having no significance what- 

If these facts be thoroughly under- 
stood and remembered, the opportuni- 
ties for misunderstanding on the part 
of the user, and misrepresentations on 
the part of the seller — which latter 

are happily of comparatively rare oc- 
currence at the present time, thanks to 
the establishment of illuminating en- 
gineering — will be reduced to a mini- 


The Boron Jewel 

It is understood that the research 
laboratory of the General Electric 
Company, which is commonly reported 
as spending $200,000 annually in its 
quest for new things of commercial 
profit, has succeeded in devising a new 
method of isolating the element 
boron in a crystalline form of abso- 
lute purity. The value of boron to 
the electrical industry lies in the fact 
that it is a non-conducting substance 
of great hardness, which can be cut 
like the diamond or sapphire into 
jewels for use in electric meters and 
other instruments of precision. 

Boron is much harder than the 
sapphire in any of its forms, all of 
which it scratches readily, being very 
close to the diamond in hardness. 
This fact has been known to chem- 
ists and mineralogists ever since 
Wohler and Deville, in 1856, at Goet- 
tingen, succeeded in converting 
amorphous brown boron into the 
crystalline form, though they were un- 
able, owing to the crudity of the 
chemical process employed by them, to 
get crystal boron - of perfect purity. 
The boron made by the laboratory at 
Schenectady is said to be the first 
which is chemically pure. It is un- 
derstood that the process of manufac- 
ture is by the electric furnace, pro- 
ducing a cheap and easily manipulated 
product, which can be molded into 
shape while in a liquid condition. Thus 
it seems possible to produce a meter 
jewel much superior to the sapphire, 
of a close hardness to the diamond, 
and at a very low cost. 

The significance of this industrial 
discovery will be understood when it 
is considered that there were about 
300,000 electric meters manufactured 
in this country last year. More than 
half of them carried two jewels per 
instrument, so that the total number of 
jewels cut from sapphire for this pur- 
pose was about 475,000 in the year. 
A jewel costs the manufacturer from 
30 to 40 cents apiece, so that we have 
here an expenditure of more than 
$150,000 annually for jewels. This 
was, however, the year of largest out- 

It may be useful to sketch briefly 
the chemical process hitherto em- 
ployed in the preparation of crystal 
boron, as it is suggestive of the means 
to be employed to reduce it in the 
electric furnace. 

Boron occurs most commonly in 
nature in the well-known form of 
borax (Ma,,B 4 O : -MoII.,0). from 

June, 1908 



which boron trioxide is readily ob- 

Ten parts of boron trioxide and six 
parts of metallic sodium are mixed in 
a crucible already heated to redness, 
and covered with a layer of powdered 
sodium chloride well dried. As soon 
as violent reaction subsides, the mass 
is stirred with an iron rod until the 
sodium is oxidized, and then carefully 
poured into water, while the boron re- 
mains behind as a brown amorphous 

This is then collected on a filter, and 
it must be carefully dried, as it is easi- 
ly oxidized and may take fire. Amor- 
phous boron does not undergo change 
in air or oxygen, nor melt at an or- 
dinary white heat, though it is easily 
fusible in an electric furnace. 

"If amorphous boron be pressed 
down tightly in a crucible, a hole 
bored in the center of the pressed 
mass and a rod of aluminum dropped 
into the hole and the crucible heated to 
whiteness, the boron dissolves in the 
molten aluminum and separates out in 
the crystalline form when the metal 
cools. The aluminum is then dis- 
solved in caustic soda, and thus the in- 
soluble boron is left in large trans- 
parent yellow or brownish crystals. 

"The same modification may be ob- 
tained by melting together boron tri- 
oxide and aluminum, forming smaller 
crystals, often joined together in long 
prismatic needles. 

"In order to prevent the action of 
air upon the fused mass, the crucible 
is placed in a larger one, filling in the 
space with powdered charcoal. 

"In this process boron takes up 
2 to 4 per cent, of carbon, prob- 
ably in the form of diamond carbon. 
There is also a certain percentage of 
iron and silicon from the crucible. 
These impurities can be removed by 
treatment with HCZ, and afterward 
with a mixture of HN0 3 and HF1. 

"These crystals of adamantine 
T>oron, according to Hampe, contain 
aluminum as well as carbon, and pos- 
sess a constant composition B 48 C,A1 3 
— Roscoe & Schorlemmer" 

Sapphire and Diamond Jewels 

For twenty years the electrical in- 
dustry has been endeavoring to discov- 
er some cheap form of jewel for sup- 
porting the moving part of meters 
and instruments of precision. Where 
the movement is only occasional, 
quartz or agate suffices very well, but 
where the movement is continual, as 
in electric meters, even the sapphire 
(hardness 9) wears away and the sup- 
porting jewel soon loses its proper 
curvature. For this reason the dia- 

mond, in spite of its cost, has been 
much employed in the commercial 
electric meter. 

Sapphire jewels are chiefly im- 
ported from Ceylon, and until lately 
have been cut on the continent at 
Geneva and at Amsterdam, costing 
about 30 cents apiece in the finished 
form, though the price is subject to 
wide variations, owing to the uncer- 
tainties of the supply. The discovery 
of boron for jewel work will free the 
manufacturer from the uneven jewel 
market and justify its adoption for 
this reason alone. 

The two most common forms of 
meter on the market are the Westing- 
house induction meter and the Thom- 
son-Houston recording meter. The 
Westinghouse meter has a very light 
moving part and the Thomson-Hous- 
ton meter has a relatively heavy mov- 
ing part, which is supported upon a 
jewel which in turn rests upon a 
spring. By long and varied experi- 
mentation, the General Electric Com- 
pany has determined that the life of 
the jewel in this form of meter de- 
pends upon the relation between the 
strength of the spring and the weight 
of the moving part, manufacturing 
several types of springs. 

The necessity for this variation in 
the spring part is due to the compara- 
tive softness of the sapphire jewel, 
which wears out quite rapidly in this 
form of meter — the average life of its 
sapphire jewel is about one year, when 
the jewel must be replaced. To main- 
tain absolute accuracy of registration 
it is necessary also to recalibrate this 
form of meter about once in every 
three months, which is a serious ques- 
tion for a central-station. It is the 
general custom in progressive operat- 
ing companies to go over the meters 
in service once a year. So far as we 
are aware, no central-station inspects 
and tests its meters as frequently as 
this interval would require. 

Much care must be exercised in in- 
stalling this form of meter in order 
to protect the jewel from damage to 
its surface, which occurs whenever the 
meter is located so that the moving 
part is subjected to jarring. For this 
reason users are cautioned not to place 
a meter upon a wooden partition, or 
other unstable structure; or upon a 
wall, which is liable to form a sound- 
ing-board, and thus magnify by 
rhythmical acceleration the ordinary os- 
cillation of the moving part, which is 
frequently sufficient to cause the 
meter to give out a slight hum in op- 
eration. In this form of meter the 
life of the sapphire jewel probably 
does not extend beyond a million revo- 
lutions of the moving part. 

In the form of jewel support manu- 
factured by the Westinghouse Co. 
there are two sapphire jewels sepa- 
rated by a hardened steel ball of 
minute size. The life of this jewel is 
found to be from three to ten years, 
and the moving part will run about 
5,000,000 rev. before either of the 
jewels must be replaced. The greater 
life of the sapphire jewel in this, form 
of meter is due, first, to the extremely 
light moving part and to the clever 
form of bearing just described. In 
the Westinghouse laboratory one of 
these meters, which has been running 
continuously for three years, has reg- 
istered 10,000,000 rev. without ap- 
preciable error, and after 25,000,000 
rev. was found to be only 4 per cent, 
out on a 2 per cent. load. This re- 
markable performance, while in a 
measure due to the type of construc- 
tion, must, since it exceeds the average 
performance, be attributed to excep- 
tional hardness of the sapphire jewels. 

It is well known that the hardness 
of the commercial form of sapphire 
bearing varies a good deal, and while 
formerly the Ceylon sapphire was ex- 
clusively used in meter construction, 
a variety of sapphire found in Mon- 
tana has lately come into extensive 
use. Its color is milky white, as dis- 
tinguished from the clear sapphire of 

For some years back diamond jewels 
have been extensively used in the heav- 
ier types of meters, particularly for 
switchboard work. The General Elec- 
tric Co. and the Edison Companies 
have experimented long in an effort to 
obtain a suitable diamond jewel for the 
commercial electric meter, but thus far 
have permitted the little information 
to become public. 

The diamond jewel is flat because 
it is well-nigh impossible to concave 
and polish the stone in a satisfactory- 
manner. It, therefore, is necessary to 
use a ring-stone of sapphire in order 
to hold the meter staff on the center 
of the diamond face. Since this form 
of jewel is practically indestructible, 
its cost, which is about two dollars 
per jewel, does not detei its use in 
meters with heavy moving elements, 
or where there is unavoidable vibra- 
tion of magnitude. It is, indeed, com- 
mon practice to provide meters of 
considerable capacity with diamond 
jewels, since their use insures accuracy 
at light load. Diamond jewel meters 
frequently show an erratic and vari- 
able speed at constant current, due to 
the fact that the meter staff frequently 
gets into contact with the ring-stone, 
thus increasing the meter friction. 

Meter Department of the Central Station 



THE introduction of motor meters 
by central-station companies to 
replace the old Edison chemi- 
cal meter* tended at first toward con- 
siderable loss of revenue from friction 
in the meters, due to the fact that after 
a short period of installation the meters 
ran slow, especially on light loads. 
The large percentage of error on light 
loads, and the imperfect understand- 
ing at the time of the causes and reme- 
dies of such inaccuracy, doubtless de- 
layed the introduction of metered serv- 
ice in place of the old flat-rate or 
contract system which was in vogue in 
the early days of electric lighting. 
Even after the meter situation began 
to be improved — on the one hand by 
improvements effected by meter manu- 
facturers, and on the other hand, by 
better care of the meters by the op- 
erating companies — the fact that the 
meters, once installed, must be main- 
tained in accurate running condition 
was slow to be recognized. 

In a catalog of the old Stanley In- 
strument Company it is stated that 
"more than 60 per cent, of the cur- 
rent consumed in electric lighting in 
the United States is passed through 
the meters at a load of less than one- 
tenth of the capacity of the meter." A 
meter's failure to register on a load 
of one or two amperes may mean a 
loss of 50 per cent, in the revenue from 
residence customers. The same condi- 
tions are, however, found in churches, 
schools and other public buildings 
(banks, libraries, theaters, office build- 
ings, etc.), where night lights, or 
watchmen's lights, are employed. 
Many central-station companies have 
recognized and endeavored to offset 
this state of affairs by establishing a 
"minimum charge" on consumers' bills, 
sometimes termed "meter rent." 

The majority of the early electric 
light companies purchased meters in 

*The writer has heard that in the early days of 
the use of the Edison Chemical Meter, it was the 
custom in some companies to estimate the cus- 
tomers' bills by a mere inspection of the amount of 
deposition of zinc on the meter cathodes. Thus the 
"expert," seated at a table on which were piled 
the zincs just removed from a group of consumers' 
meters, would pick up each zinc in turn, examine it 
critically, and announce "No. 24, $1.40; No. 47, 
$3.00; No. 10, S2.25"— and so on. This practice 
has, of course, sooner or later superseded by the 
system of accurately determining the increa 
weight of the cathodes; a system which remained 
in operation, subject to various refinements and 
modifications, until the discontinuing of the use of 
the Edison meter. 


a desultory way — the "Meter Depart- 
ment," in the rare instances in which 
.that term was used, having an irre- 
sponsible, or, at any rate, an indefinite 
or experimental status — and installed 
them in an equally desultory way, with 
a childlike reliance upon the claims of 
accuracy made by the meter manufac- 
turer whose product happened to be 
preferred. The matter generally ended 
there, so far as any consistent or sus- 
tained system of meter maintenance 
was concerned. Even after the use of 
motor meters at consumers' installa- 
tions had entirely replaced both the 
old flat-rate system of charging — the 
charge being usually made on a "per 
i6-c-p. lamp per month" basis- — and 
the employment of the Edison chemical 
meter, the central-station companies 
were remarkably slow in realizing the 
importance of efficient organization in 
connection with the device upon which 
their entire revenue depended. Up to 
a recent date many of the best man- 
aged companies, with elaborate and 
costly layouts for securing maximum 
economy in the generation and dis- 
tribution of electricity, have been slow 
to sanction expenditures to secure 
efficiency at the other end of the propo- 
sition — the actual metering of the 
product as delivered to the consumers. 
Oftentimes the utmost care has been 
exercised in preventing waste at the 
coal pile and in securing efficient boil- 
ers and engines, and money has been 
spent cheerfully for the installation of 
larger and more efficient generators 
and the laying of heavier conductors, 
while the meters have been allowed 
to take care of themselves. 

Though a meter that "runs fast" 
may occasionally be found, due to ex- 
traordinary conditions or accidentally 
defective adjustment, the natural ten- 
dency of a meter, as of any machine, 
is to "run slow" in service. Adequate 
calibration, therefore, much more than 
pays for itself in preventing losses 
through underregistration. 

The advisibility of maintaining the 
accuracy of a central-station com- 
pany's meters by adequate rating and 
calibration is demonstrated by the fact 
that a I per cent, increase in revenue, 
secured by such maintenance is 
equivalent to a saving of several per 
cent, in the coal bill. Moreover, only 
by maintaining- the accuracy of its 
meters can a company be in a position 

to adjust its charge equitably for all 
of its customers. Occasional cus- 
tomers whose meters are found to be 
registering far too low may be dis- 
gruntled on finding that their meters 
are running faster following the visit 
of the inspector, and may order the 
service discontinued. Such customers 
are, however, undesirable anyway, 
since acceding to their demands re- 
sults either in loss of the power de- 
livered to them or in injustice to other 
consumers by the effect on the rates 
of such virtual discrimination. 

Modern methods demand high 
efficiency in all departments, however, 
and with growing realization, by cen- 
tral-station companies, of the im- 
portance of the meter proposition, it 
has become well recognized that in ad- 
dition to selecting correctly designed 
and well-constructed meters and in- 
stalling them with care, the meters 
must be systematically maintained at 
their initial accuracy. 

In this connection Mr. \Y. J. Mow- 
bray, in a paper entitled "Maintenance 
of Meters," presented at the A. I. E. 
E. meeting of April 28, 1905, says : 

"To the company supplying electric 
energy which is measured by meters 
and charged for accordingly, the main- 
tenance of meter accuracy is of su- 
preme importance. Losses in other 
apparatus become insignificant when 
compared with the loss of revenue 
from meters that are allowed to fol- 
low their natural tendency to run slow. 
For example, in a steam boiler a drop 
of 10 per cent, from normal efficiency 
would be detrimental to approximately 
the same percentage in the single item 
of the cost of coal, whereas in the 
meter system it would be 10 per cent, 
of the entire gross revenue to which 
the supplying company is legitimately 
entitled. Furthermore, if a metering 
system did actually deteriorate so as 
to record 10 per cent, less than the 
true energy, this loss would by no 
means remain constant ; it would con- 
tinue to increase. 

"Periodic overhauling is the obvious 
and generally adopted means of main- 
taining meter accuracy. Overhauling 
— a strict examination for correction 
and repairs — is efficient in proportion 
to the cheapness and accuracy with 
which it is done, and to the per- 
manence of the result." 

In recognition of modern ideas, 

June, 1908 



the meter departments of central- 
station companies have been given 
more scope. The increased efficiency. 
showing as direct increase of revenue, 
to be attained by consistent testing of 
the connected meters, and the good re- 
sults to be attained by handling the 
clerical work connected with in- 
stalling, reading and testing of the 
meters directly by. or in close co- 
operation with, the meter department 
have been recognized, and in the 
larger companies have led to some- 
thing like the following organization : 

1. A stockroom and construction 
department for storing, installing and 
removing consumers' meters. 

2. A testroom for the systematic 
testing and calibrating of the meters 
before installation, and after removal 
for any cause from consumers' 
premises ; with facilities for repairing 
and making over meters, and a sys- 
tem of routine and special testing of 
the meters on the customers' premises. 

3. A combined meter-reading and 
bill-computing organization, handling 
all routine and special meter reading 
and the entire clerical work connected 
with the actual billing of the con- 

These different parts of the meter 
work are discussed in other chapters. 

Although the meter departments of 
central-station companies do not 
usually embrace the management of 
all the above divisions of the meter 
work, it is recognized that there 
should be free and intimate co-opera- 
tion among them. 

The importance of such co-opera- 
tion for the good of the service may 
be illustrated, for example, by the 
need of agreement between divisions 
1 and 3 on the location of meters 
to be installed, involving consultation 
in some individual cases to ensure that 
the meter is located to the best advant- 
age for testing and reading. Again, 
division 2 and 3 need to keep in touch 
to ensure the laying of concordant 
(non-clashing) meter routes for test- 
ing and reading, respectively. 

In a few companies, however, the 
advantages gained by close co-opera- 
tion and a common clerical system 
have led to the including of all the 
above divisions in the province of a 
single Meter Department. 


The accurate reading of installed 
meters is obviously a matter of great 
importance. Errors in reading con- 
sumer's meters result in inaccurate 
bills, which are exasperating no less to 
the company than to the consumer, 
and it is always advisable to employ 
as meter readers only men who are 
thoroughly competent and trust- 

worthy. With many companies it is 
customary to shift the men about on 
the different "meter routes" or sec- 
tions, so that no one man reads the 
same route twice consecutively. 

The meter readers are often in- 
structed to make a superficial ex- 
amination of the meter and its wiring. 
etc., and to report anything defective 
or irregular about the installation. 

In most companies the meters are 
usually read once a month, except in 
special cases, where it is necessarv t< 1 
render bills weekly. In small and 
medium-sized systems, where a com- 
paratively small number of meters are 
in use, it is customary to read all of 

ds or pointers on a printed. fac- 
simile of the meter dials. 

Many different styles of printed 
forms for recording the dial readings 
are employed. Fig. 1 shows a form 
printed on the back of a "meter card." 
designed for one year's readings. 
These cards are filed by meter routes, 
i. c, in the order in which the meters 
are to be read. The cards for any 
meter- disconnected can thus be re- 
moved, and the cards for meter- added 
can be inserted in their proper places. 
< )ne such diagram is provided for each 
reading of each consumer's meter, and 
the meter reader marks upon it the 
p isition of the pointers and turns it 


the meters during the last few days of 
the month and present the bills on or 
about the first of the following month. 
On large systems in which the great 
number of meters renders this method 
impracticable, it is customary to divide 
the city into sections and read one 
section every day. 

Methods of Recording the Dial 
Readings. The following two general 
methods of recording the readings are 
in use : 

1. By marking the positions of the 

in as a permanent record, to be re- 
duced to watt-hours by some responsi- 
ble person. The chief advantage 
claimed for this method is that it in- 
sures closer inspection of the dials by 
the meter reader, as he is required 
to record the positions of the hands 
to represent as nearly as possible their 
true positions on the dial. The rec- 
ords also furnish a complete history 
of the successive positions of the dial 
hands, month by month, which, in the 
case of complaint resulting from 



June, 1908 

loose or misplaced dial hands, is fre- 
quently o.f service in adjusting the 

As . against these claimed advan- 
tages, it is fair to assume that ability 
to mark a dummy card correctly for 
all positions of the dial hands should 
qualify the meter reader to record the 
reading directly in figures : a proposi- 
tion which implies that the method is 
more useful for training the novice 
than for regular use by the expert. 

2. By recording the reading directly 
in figures ; the advantage claimed for 
this method being that it is some- 
what more rapid and less cumbersome. 
It is believed that most expert meter 
Teaders employ this method. 

Directions for Reading the Dials. 
Whichever of the two methods is used 
certain positions of the pointers are 
difficult of interpretation, necessitating 
great care in reading the register. 
Hence the following directions should 
be carefully observed : 

A. Note carefuly the unit in which 
the dials read, i. e., whether in watt- 
hours or in kilowatt-hours. In some 
makes of meters (e. g., General Elec- 
tric Company meters) the number 
marked above or below each dial in- 
dicates the value of one complete revo- 
lution of the corresponding pointer, so 
that one division on the dial indicates 
•one-tenth of the printed number. In 
other makes of meters (e. g., the West- 
inghouse) the printed number indi- 
cates the value of one division on the 

B. Note directions of rotation of 
the pointers. Thus in General Electric 
meters the pointers of the first, third 
and fifth dials (counting from the 
right) rotate in the clockwise direc- 
tion, whereas the pointers of the sec- 
ond and fourth dials rotate in the 
counter-clockwise direction. 

C. Read the dials the reverse in 
'order of their value, i. e., beginning 
-with the right-hand dial. 

Most expert meter readers, how- 
ever, read the dials in the'direct order 
■of their value, i. e., in the order in 
which the figures are set down to ex- 
press the reading in kilowatt-hours — 
except in the case of very difficult 
readings, which have to be "built up" 
carefully from right to left. 

D. Always read the figure on each 
dial which the pointer has last passed 
over or which it just covers. But note 
carefully that the reading of each dial 
depends upon the reading of the one 
next to it on the right. Unless the 
next pointer on the right has com- 
pleted a revolution, as shown by its po- 
sition directly over the zero or a little 
"way past it, the pointer which is being 
read has not completed the division in- 
dicated by the figure upon which it 
may appear to rest, and still indicates 
the figure last passed over. 

E. See whether the register is direct 
reading, or whether it has a multiply- 
ing constant. Some registers are not 
direct reading, but require that the 
reading be multiplied by a constant in 
order to obtain the true number of 
watt-hours. Under no circumstances 
should a constant be used in record- 
ing a reading unless the dial face bears 
a multiplying factor ; for example, the 
words "multiply by l / 2 ," or "multiply 
by 10." 

The necessity of exercising care in 
reading the dials is illustrated by the 

Nos. 2, 3, 9, and io may be taken as 
examples of "difficult readings." In 
any meter register the pointers some- 
times become slightly displaced, as 
shown in Nos. 3 and 4, but the actual 
indication may be determined by the 
reading of the pointer at the right of 
the one in question. 

Additional examples of dial read- 
ings are given in ' the following ex- 
planatory text referring to Fig. 3, 
with directions for billing reproduced 
from a publication of the Duncan 
Electric Mfg. Company. The registers 

1" uoaao 

:: Ml 

eooo A 



No. 1 






No. 6 


tat ^ 

\ .-03 090 


/M ^W*\ 

No. 2 

■* 5- 



. % 

No. T 


, '"^ 



. * 







No. 3 

V v 




i\(>. 8 









<%k *" 




No. i 




' * 

No. 9 














No. 5 





No. lu 





accompanying examples of positions of 
the dial hands on General Electric 
registers, Fig. 2, in which, as already 
stated, the number marked near each 
dial indicates the value of the entire 
revolution on the dial. 

The correct readings for the ten 
registers shown are as follows : 

No. 1 1,111,100 

2 999.900 

3 1,000,100 

4 9.999.500 

5 909,100 

6 99.7oo 

7 9,912,100 

8 9,928,000 

9 9,918,100 

10 9,928,300 

shown are of the standard Duncan 
type in which the number marked 
over each dial denotes the value of 
each division on the dial : 

"The values (1000s, 100s, 10s, is, 
Tenths) over the dial circles refer to 
the divisions of the circle over which 
they stand. 

"Therefore, a division on the dial 
circle to the extreme right indicates 
one, two, three or four tenths of a 
kilowatt-hour, while a complete revo- 
lution of the hand or pointer would 
be 10 tenths or one kilowatt-hour, 
and will have moved the pointer on 
the second dial circle one division 
(one kilowatt-hour). 

"Thus in reading dial No. 1, the 

Jtme, 1908 



first dial circle (that on the extreme 
right) indicates i (one- tenth) the 
next (is) indicates I, the next (ios) 
indicates I, the next (ioos) indicates 
1, and the remaining dial circle 
(iooos) also indicates I, making the 
total reading or indication I I i I .1 

"A hand or pointer to be read as 
having completed the division must be 
confirmed by the dial before it (to 
the right). It has not completed the 
division on which it may appear to 

No.1- 1111.1 

then the reading will be iooo kilowatt- 

"The hands are sometimes slightly 
misplaced. In dial No. 8 the first dial 
circle (the extreme right) reads o 
(no tenths). The hand of the second 
dial is misplaced. As the first registers 
o, the second should rest exactly on 
a division ; therefore, it should have 
reached 8. The three remaining dials 
are correct and make a total of 9928.0 

"In dial No. 9 the second dial hand 

N0.6 -99.7 


»oo* 10* 



No.7- 9912.1 



kilowatt hours. 

0O 3 iO* I s * 



fs Sw 


I eVs 1 1 

i 2 i 8 

r 8 » *i 

V 3J 


7 A 7 •*; 

V 7 1 
\4 5 6/ 

v V 




-O. W*TtTT( ,^ct USA 


«0O4 10* 




too* iO» i» 




ZY2 I sYs I 2 V2 1 

sAs 'A' 'A' 



a (fiS 
■ 7 zj 

^ OJ«<C»s Z 






iOO» iO» I s -tt**TN 3 


fa 1 


vf 2 A 8 Y 8 a»-?T 2 ©-*T 8 I 2 ! 

^ S.^£>-. £l 

Duncan integrating wATTMtTCR- 

tC«IC MA-NuFACToniMG CO. L»» r *'«E'TJ iND. 1/ J. 


No. 4-9999.4 







lOO> IO s I i 


f 8 a 

I7 • 

\? 5 ' 

13 7^7 lis ir 7, 

"* rf 2 1 

< 7 *7 

^ 3^^C*s E 



TAlC'.C'uWIia CO.lA#*V(TTe ,.O.v5. 


No. 5- 909.1 

v ooc 

V »«~«A 

vd lO-va> t hOURS ^ 
J ,00> iO» i 1 4U*THs 

§ft /;/• /5Y5 f%» Si 

OunCAs a,'cgo«.t,ng AATTMETER 




rest, unless the hand before it has 
reached or passed o, or, in other 
words, completed a revolution. There- 
fore, it is always advisable to read 
dials from right to left. 

"In reading dial No. 2, the first dial 
circle (to the extreme right) indicates 
■0.9 (nine-tenths). The second hand 
apparently rests on o, but since the 
first rests only on 0.9 and has not yet 
completed its revolution, the second 
<lial circle also indicates 9. This 9, 
placed before the 0.9 already ob- 
tained, gives 9.9. This is also true of 
the third dial circle. The second dial 
circle hand at 9 has not yet com- 
pleted its revolution, so the third has 
not completed its division; therefore, 
another 9 is obtained, making 99.9. 
The same is true of dial circle four, 
thereby making the total reading 
999.9 kilowatt-hours. When the hand 
on the first dial circle (extreme right) 
-completes its revolution or reaches O, 

is misplaced, for since the first indi- 
cates 0.1 (one-tenth) the second 
should have just passed a division. 
As it is near to 7 it should have just 
passed that figure. The remaining 
three dial circles are approximately 
correct. The total indication is 9918. 1 

"In dial No. 10, the second dial 
circle hand is slightly misplaced by 
being behind its correct position, but 
not enough to mislead in reading. The 
total indication is 9928.3 kilowatt- 

"By carefully following these direc- 
tions little difficulty will be experi- 
enced in reading the dials, even when 
the hands or pointers become slightly 

"These dials read direct in kilowatt- 
hours (thousands of watt-hours), and 
as this is the unit upon which the 
rate of charge is based, it is obvi- 

ous that it should meet with approval 
from the central-station managers. 

"The following is an example of 
making out a bill on kilowatt-hour 
basis : Suppose the dial reading is 
21.8 kilowatt-hours at 20 cents per 
kilowatt-hour or per 1000 watt-hours, 
which is the same thing, the amount 
will be 21.8 X 20 cents = $4.36. If 
the rate is 16 cents the amount will be 
21.8 X 16 cents ± $3.48. If the rate 
is 10 cents the amount of the bill will 
be 21.8 X 10 = $2.18. 

"If the dial has 'multiply by io' 
marked on it, the reading must be mul- 
tiplied by 10. Example : Dial reading 
46.8 kilowatt-hours multiplied by 10 
equals 468 kilowatt-hours." 

The increase in the number of elec- 
tric meters installed, on both old and 
new consumers' installations, has led 
to more or less systematic attempts to 
educate the public in reading meters 
and computing "bills for current," and 
some of the meter manufacturing and 
central-station companies have issued 
brief instruction cards or booklets on 
the subject couched in simple and 
non-technical language. 

Two such publications are a small 
eight-page illustrated pamphlet en- 
titled, "How to Read Your Electric 
Meter," issued by the General Electric 
Company (No. 3523, December, 1906), 
and a small 12-page illustrated pamph- 
let entitled, "Wattmeters and How to 
Read Them," issued by the Westing- 
house Company (Folder 4032, July, 
1907). Another and more extensive 
publication is a 56-page illustrated 
pamphlet entitled, "How to Check 
Electricity Bills," by S. W. Borden, 
published by the McGraw Publishing 


In many central-station companies 
the factory numbers of the meters 
owned by the company are used in the 
records of the meter department. 
But where a large number of meters 
are in use by a company, it has been 
found desirable to renumber the 
meters in a consecutive series, going 
by the order in which they are bought. 
Thus a company owning a number of 
meters of different makes (the factory 
numbers of which are not, of course, 
consecutive) will apply a new number 
to each meter in a consecutive series 
from number 1 up, irrespective of the 
make, type or capacity of the meter. 

While this renumbering plan facili- 
tates in some degree the keeping of 
meter records, etc., as compared with 
dependence on the factory numbers of 
the meters only, the practice is grow- 
ing of marking each meter on a num- 
bering system designed to express at 
a glance not only the company's serial 
number of the meter, but also the 
make, type and capacity of the meter. 



June, 1908 

Each meter number under this system 
is a composite number consisting of 
two parts separated by a symbol or 
letter referred to a code and designat- 
ing the make of the meter. The first 
part is the serial number of the meter, 
in a series running from one up, for 
each make of meter owned by the 
company. The second part, called the 
"capacity number," consists of an ar- 
rangement of three digits referred to 
a code, and expressing in order the 
size, voltage and "wire" of the meter, 

The following codes, which are 
used by certain central-station com- 
panies under this system for designat- 
ing the symbols, and the three digits 
of the "capacity number," will serve 
to make the general method clear. 


Make of Meter • Symbol 

TRW Commutator Type — 

TRW Induction Type, Single Phase + 

TRW Induction Type, Polyphase P 

Stanley Recording Wattmeter S 

Fort Wayne Integrating Wattmeter F 

Size, Digit 

Amperes (Number or Letter)* 

3 1 

3H 2 

5 3 • 

1V2 f 

10 5 

15 6 

25 7 

50 8 

75 9 

100 A 

150 B 

200 C 

300 D 

♦Letters are used for meters of capacities of over 
75 amp., to avoid the use of two digits to designate 
these capacities. 









The following table gives examples 
of this system of numbering : 

Meter No. 

1120 - 523T R W 

15 + 112T R W 

246 P 843T R W 

872 S 612Stanley 
1064 - B52T R W 

Description of Meter 

Commutator Type, 10-amp., 

220 volt., three-wire. 
Induction Type Single-Phase, 

3-amp., 110 volt, two-wire. 
Induction Type Polyphase, 50- 

amp., 440 volt, three-wire. 
15-amp., 110 volt, two-wire. 
Commutator Type, 150-amp., 

550 volt, two-wire. 

In some companies the "capacity 
number" is given first, and the "serial 
number" second ; instead of the other 
way about as above described — the 
general scheme being otherwise un- 
changed. For meters having room on 
the cover, free from other markings, 
etc., the number may be painted or 
stenciled on — white-lead paint and a 
neat stencil 'gives good results — or 
may be embossed by machine on an 
aluminum plate which is then riveted 
to the cover. 

The disk constant of the meter is 


Serial No Factory No District, 

Route Date , 

Customer's Na^e , 


Premises Occupied as 

Meter Location: Put up on... 


Occupied as, 

Meter Typo Form Cat. No, 

Capacity Amp. Constants: Dial 

Volts Test 

Wire Bill 

Potential Volts Soal Found , 

Circuit Wire Volta at Meter Phase_ 


Load: Lamps 8 c.p 16 c.p &2 c. p.. Special 

Motors H.P Fans Arcs 


Wattmeter No Voltmeter No Ammeter No 

Milli-Voltnotor No Shunt No. . . .Stop Watch No. .Seadins Tool I'o. . . 

Time Arrivyd 

Ah Pound 

Per cent of; 
Full Load 




'Rov, :s©c f (is 

In s t . 

■ Watts 

Per C?nt ; 
: Accuracy : 





1 ' 25 




100 : 

Time Left As Left 



3 5 



Lamps Required to start, ha fore a dJ After Ad j , 






















Test by 


Last Test: Page No. 


NO I. 

June, 1908 



sometimes put on the cover also, 
where it forms a valuable part of the 
meter description which is of special 
assistance to the meter inspectors. 

Among the advantages of this sys- 
tem of meter numbering may be men- 
tioned the concise presentation of the 
information needed to identify the 
meter independently of other meters, 
and the assistance afforded to meter 
inspectors, to the stockkeeper and to 
the accounting department. The em- 
ployes of the meter department soon 
accustom themselves to the system so 
as to 'identify the meter at a glance 
without removing the cover. The sys- 
tem is of special advantage in .the 
stock room, as the stockkeeper can 
always tell quickly — as when deter- 
mining on a requisition for new me- 
ters, or taking an inventory — the num- 
ber of meters of all makes and sizes 
that are on his racks. 


In small companies the stock of 
meters may be kept on shelves in the 
common room appropriated to meter 
work. In many large companies a 
separate room is set aside as a stock- 
room and fitted with shelves and open 
racks to hold the meters, with lockers 
for meter parts, tools and supplies, 
and often for the company's equip- 
ment of testing instruments. As 
stated above, the orderly keeping of 
meter stock is greatly facilitated by the 
use of a serial numbering system for 
the company's meters. 

The shelves and lockers should be 
kept clean and neatly painted, and the 
stock — which of course is constantly 
changing — should be arranged so that 
the stockkeeper always knows how it 
stands. The ideal should be the keep- 
ing of the stock in clean shape, with 
a minimum number of idle, inopera- 
tive and out-of-date meters on the 
shelves or lying about on the benches 
of the repair shop, yet in such condi- 
tion that a rush requisition may al- 
ways be filled. 

The stock record generally takes 
the form of a card index, as described 
under Office Routine. 


The clerical management and records 
of meter stock, and of the installa- 
tion, regular and special reading, and 
regular and special testing of meters, 
are handled in different ways, accord- 
ing to size and organization of the in- 
dividual companies and the prefer- 
ences of their officials. The brief re- 
marks herein have been prepared to 
indicate merely the general scope of 
this routine and a few of the actual 
methods that are employed by differ- 
ent central-station companies. 

Printed blanks are employed for the 

meter readers' and inspectors' 
use, in the form of a book, loose slips 
or cards. 

The following are typical forms of 
printed pages of meter inspectors' 
books, each laid out for a single test : 

The form marked No. i contains 
spaces for test observations only. In 
order to simplify the record of the in- 
spectors' work upon the meter, code 
letters are provided under "Adj." (ad- 

tions made with the test meter (T. 
M.). It is made up with alternative 
entries, and the inspector is instructed 
to check what he has found or had to 
do in the meter, and to cross off what 
he has not found or had to do. This 
particular record shows that the dial 
constant is 2, that the inspector who 
last opened the meter was Xo. 6 
(seal), and that the dial mechanism 
was tested to make sure that the gear- 

dYc., penVtfic, 

1 pliii jpeiljl. 
3 ph^ i m p 

Meter Test date 

JJ-0O /-</ Ct&'u-'r^fc'n < 




/^"amp 230mi$ 3 wire 





DIAL Gift.. DISK K ..——. TOP BEARING -high, lovfr-K. 

CREEP "?"'iniinV. occjlfonal, RATE .- .- 

DUE TO 7 = k<&dA*%^ A>£^^ ■m—/&-~~ RRSTPtt-flAY 

As Found 

JEWEL tyr oipphirfl, y 


(t/iis section fur inc 


icating instruments) 


Correct Actual 





of actual energy record- 
ed by consumer's meter 

_ Correct rev. of T. M. 



Actual rev. of T. M. 

. /.*r 







> Lef 1 














/0O if 




/( r o . 






*K- cup dWmonrJ. 
JEWEL nfr»rN j|>p[i lii nji 

ADJUSTED najrfet. shunWo'il, top oVatlng, 

CLEANED -ff— mutitnr hrnrhnii inii^lli I 

BRUSH TENSION \ ' E GH J JTffj; ^"S fl 

m.lV./}l<rVLly SEAL NO. K 

f hj l II ,l l C lI IIM Lll i u il i un6 Jfll lll ul g 

HOW TES1 ED —J oo nn ooti a n , two amm e t e rs o r tw o t a rt mot e rt , fieldsAtfriesd. _ 

INSTRUMENT NO /^>~2- TIME OF W£t../0.~T TO .10 fe~ 




'Ifit unnecessary to make diagram, remarks, etc., reverse carbon and write on back 
of yellow sheet. 

NO 2. 

justments), referring to the different 
parts of- A. C. and D. C. meters, and 
the inspector indicates the parts which 
he has had to adjust by checking with 
his pencil the proper letter. 

Xo. 2, which is a filled-in form, 
contains space for the dial reading a- 
well as for the test observations, and 
is designed to reduce as far as pos- 
sible the number of entries, i. c. . the 
amount of writing required of the 
meter inspectors. The form is laid 
out primarily for recording observa- 

ing connecting with the meter shaft 
was "in mesh," and whether all the 
wheels and hands were tight on their 
individual shafts. There was no con- 
stant i K i marked on the disk, the 
meter being of the style having the disc 
constant the same as the dial constant. 
The top hearing was found "low," 
which should indicate to the tester that 
it should be raised (after As Found 
test ). The meter crept occasionally, 
as stated, and a "clip" creep retrmkr 
had to be installed on the disk. T!k> 



June, 1908 

tension of the left-hand brush was 
found to be too light and was in- 
creased. The jewel was reported — 
of course, upon examination after the 
As Found figures were obtained — to 
be a sapphire and in bad condition. 
The remaining entries will be clear 
upon examination. The columns in 
which no entries appear are provided 
for the use of the form in testing with 
indicating instruments. These forms 
are furnished to the inspectors in 
manifold books consisting of 50 white 
leaves, perforated for ready removal 
and constituting the originals, alter- 
nating with 50 colored leaves bound 
securely to give the carbon copies, and 
numbered consecutively from 1 to 50. 
This scheme has the advantage that 
it calls for some marking of each and 


mounted on stubs and fastened to- 
gether in books. The larger card 
(Fig. 5) contains the meter depart- 
ment record of changes in service and 
adjustment, and gives the complete 
service history of the meter, including 
dates of receipt and installation, length 
of service, idle periods in the stock- 
room or repair shop, and performance 
in service. It has space for six en- 
tries, which enable the single card to 
cover a period of three to six years, 
according to the frequency with which 
changes are made in adjustment and 

These original records are ar- 
ranged to be filed at "the office of the 
meter department. The tendency is 
toward the use of index cards of 
standard sizes rather than a book, or 

at 1 * 

FROM) /^ 



* >^? *r 


2/s-/ 7o 























CARD Na. 16012. 

METER m.l/l'/tG M1t.WN.3h/4f MAKE IfrZW (Q (L TYPE Z^C^ . 
MFR-S. NO. j ./ ff'/ffO AMP. /Q ' VOLTS /Q Q 






/aa- & 

/ &>'*■ 

**±u jjumJJ- J y 4 w — 





every item, and any omission is read- 
ily detected by the inspector himself 
by a look over the form before leaving 
the customer's premises. 

Another record system is shown in 
Figs. 4 and 5, which are facsimiles of 
meter-record cards furnished by the 
Westinghouse Electric & Mfg. Com- 
pany. Fig. 4 shows card of the stand- 
ard three by five-inch size, which is 
filled in by the inspector at the con- 
sumer's installation. The cards are 

loose slips, especially in the larger 
companies, on account of the recog- 
nized 'advantages of the card-index 
system of filing. 

The following general clerical rou- 
tine is followed in some large com- 
panies : When the contract notice, 
duly approved, comes down to the 
meter department from the general 
office, an addressograph stencil is pre- 
pared, containing the customer's com- 
plete address. This stencil is used 

primarily for billing, but also to head 
the card containing the consumer's 
monthly and special dial readings 
("meter-record card"), and to head 
other cards for insertion in other in- 
dexes. The keeping of the meter- 
record cards in order in the files of the 
clerical section may be facilitated 
by indenting the edges of the 
card with a punch. The card of a 
given district and meter-inspection 
route are all indented at the same dis- 
tance from the top of the card, so that 
in a group of cards for a given route, 
the indentations come in line when 
the cards are racked evenly. Thus a 
mistake in filing any card with cards 
of another group shows plainly by 
the offset of the misplaced card's in- 
dentation. The contract notice is then 
turned over to the construction de- 
partment as an order to install the 
meter, and when the installation has 
been made comes back with the instal- 
lation order attached and the number 
of the meter written on the back. 

The meter readers transcribe the 
dial readings to the meter-record 
cards. They turn in a daily report, 
with the meter-record cards for the 
day. Complaints are transmitted via 
the district office. 

The practice of another large com- 
pany in 1905 is given in the following 
quotation from an article entitled, 
'"The Meter Department of the New 
York Edison Company," in the Elec- 
trical World and Engineer, April 1, 
1905: "After this test (the test that 
is made upon receipt of meter from 
manufacturer) has been completed 
and proper entry made in the record 
of the meter department, the meter 
is again returned to the storeroom to 
await installation in a consumer's 
premises. The contract and inspection 
departments being conversant with the 
demand of the customer, assigns the 
meter size and its location, and ad- 
vises the distribution department when 
the installation has been approved by 
the Board of Fire Underwriters. The 
meters are installed by the distribution 
department, a meter of proper capac- 
ity being obtained from the storeroom, 
and the latter notifies the auditing de- 
partment of the withdrawal of the 
meter from stock. When the meter 
has been installed, the contract and 
inspection department is notified by 
the distribution department. The 
former then advises the auditing de- 
partment that the meter has been offi- 
cially connected, and a new entry is 
then made for billing purposes in the 
records of the latter department. 
Prompt advice of this installation and 
connection is made to the meter de- 
partment, which is thereby enabled to 
make a regular inspection of the meter 
within one week after it has been 

June, 1908 



set." A so-called "first test" is made by route and in order of inspection in means the total number of every size 

at the expiration of four to six weeks, each route. The card (of which Fig. 
to give the (D.C.) meter's commuta- 7 is an example) contains meter de- 

and on order can readily be deter- 
mined within a few minutes. 

tor time to attain its normal service 
condition of oxidation, thereby safe- 
guarding the customer against error 
in his meter. 

"Subsequent to this first test, 
each meter is tested at least annually. 
* * * Tests are also made at the 

scription, serial number and date of 
installation, name and address of cus- 
tomer, etc., and different colored cards 
are used according to the service — 
whether A. C, D. C, 500-volt D. C, 

Numerical record of meters owned 

expense of the company upon com- by the company. These cards (Fig. 8) 
plaint of the customer that his bills are kept in groups or, in larger corn- 
appear excessive. The policy of the panies, in separate cabinets, contain- 
New York Edison Company has been ing records of meters respectively in 
very broad in this matter. If the cus- service (with customer's name and ad- 



Meter No. Make Type Form 

Amps. Volts Wire K 








Potomac Electric Power Co. 


Uo. _ i fruit-. 




Date „. Reading Signed 



tomer is not satisfied with the test dress ordered, and in stock, and are 
made by the company's representa- arranged by make, capacity and serial 

tives he is privileged to have his own 
expert present to witness the test and 
calibration. The results of the tests 
are reported on form test cards, which 
are checked by the foreman of the dis- 
trict, and proper entry is then made 
in the records of the meter depart- 
ment, so that the tests of the meters 
are available at all times, and may be 
referred to in the files of the com- 

Whether loose slips, books or cards 
are used, however, the routine should 
accomplish the following two objects: 

1. A filing system, arranged either 
by meter numbers or names of con- 
sumers or by both (duplicate records), 
that will facilitate ready reference to 
the complete history of a given meter, 
including all previous tests, or to the 
history of the meter or meters of a 
given consumer. 

In addition to "meter-record cards," 
the card-index records of the clerical 
section may include the following, 
among others : 

Installation and removal record : 
Card (Fig. 6) contains on one side a 
form for very brief data, including 
name and address of customer, dial 
reading and date, under the heading 
"Installed," as shown; and on the 
other side the same form under the 
heading "Cut off." 

number. Cards for meters of the 

same capacity may be filed together card for this record is "shown in Figs 


MtUrHc. 1 'jmlum . .,.. 


Alphabetical record by customer's 
names: The cards are duplicates of 
"meters in service" cards referred to 
in the preceding paragraph. 

Meter inspection record : Cards ar- 
ranged in order of inspection, by 
route, and containing customers' 
names and addresses with meter de- 
scription, data of load, and spaces* for 
complete test data under the heads 
"As Found" and "As Left" and for 
Remarks." An example of a double 





pivot mm tfat 







v, «' T ? 




V. a — S watts 









TEM •■ 


and make of meter in service, in stock 

serially for each make of meter, and 

groups covering the different capaci- 

Route index record cards, arranged ties separated by guide-cards. By this 

9 and 10. front and back of card, re- 
spectively ; the back of the card being 
used for special inspections on com- 
plaints, etc. 



June, 1908 

Record of diamond-jewel installa- 
tion, giving meter description^ cus- 
tomers' namo and addresses, data of 
load, date at which the original sap- 
phire jewel was replaced by a dia- 
mond jewel, and data constituting the 
historv of the diamond in service. 
Card is shown in Fig. II. 




'^S T . 


PFM. , PEA^IN". 




2. A record of each meter reader's 
and meter inspector's work for deter- 
mining the relative efficiency of the 
men and difficulty of the several me- 
ter-reading and meter-testing routes. 

A^ an illustration of what may be 
done in this direction with a minimum 
of clerical work, the following routine. 
employing the manifold-book form 
Xo. 2 described above. is in use with 
1- results in at least one rather 
large company : 

At the close of each day the in- 
spector tears off the original (white) 
leaves and turns them in as the re- 
port of his day's work. These leaves, 
after being looked over by a person 
having that duty, are tiled for future 
ready reference. The inspector re- 
tains the book with the colored leaves 
until 50 tests have been made and then 
turns it in. As each inspector's book, 
bearing his name on the outside, is 
filled in. it is given a number and 
filed in a pigeon-hole cabinet like that 
shown in Fig. 12, with a section or 
vertical row of pigeonholes for each 
meter inspector. Since each book- 
contains 50 tests, the number of tests 
to the credit of each man i- seen at 
a glance — in the illustration 400. 350 
and 500, respectively — thus stimula- 
ting a healthy competition or rivalry 
among the several men. It is a good 
idea to shift the men around in the 
various districts so as to avoid any 
claim that a certain district is much 
more difficult than others. In practice, 
the pigeon-hole case might better be 
made to hold 20 books, or 1000 tests. 
On a given date, when one section is 
full, the books arc removed to the 
record storage. 

Bill Computing. 

Billing of consumers from the orig- 
inal records of the meter readers — 
either the marked facsimiles of the 

meter dials or the records in actual fig- 
ure- already described — or from meter- 
record cards containing data tran- 
scribed from the original records, is 
performed by clerks at the office of the 
meter department or in the account- 
ing department at the general office, 
according to the organization of the 

In some large companies the con- 
sumer's bills are made out from a 

ting the making out of bills from me- 
ter readings. The chart is redrawn, 
with changes, from a large chart fur- 
nished by the ( ieneral Electric Com- 
pany. The figures at the bottom are 
kilowatt-hours and those at the left 
are the amounts of bills in dollars and 
cents. The diagonals are different 
rates per kilowatt-hour, ranging from 
5 cents to 30 cents. 

Selecting the diagonal having the 














"consumer- meter-record sheet," 
which is filled in from data on meter- 
record cards. This sheet also con- 
tains columns for "net readings." com- 
puted either in the clerical section of 

rale at which charges are to be made, 
a point is found on it directly over the 
number of kilowatt-hours shown by 
the meter. In the column at the left, 
on a horizontal line from the point re- 

"> J " -. ■*" ■*» •*• *' 7 


the meter department or in the gen- 
eral office, and is practically a ledger 
sheet, with one line for every (month- 
ly) reading. 

Fig. 13 is a price-chart for facilita- 

ferred to, will be found the amount of 
the bill. For example, if the meter 
readings show a consumption of 50 
kw-hr., and the rate is 10 cents per 
kw-hr., the amount of the bill is $5. 

Cable Insulation 

Compiled from Notes by "W. -A.. Del Mar 


THE principal materials used for 
insulating power cables are : 
I i i Paper saturated with oil. 

(2) Varnished muslin, variously 
known as varnished cambric or var- 
nished cloth. 

(3) Compounds containing rubber. 
The first two being made of staple 

commercial materials are generally re- 
liable, but compounds containing rub- 
ber vary from the cheap material used 
for insulating "code wire" to the high- 
grade compound required by the U. S. 


If two conductors are arranged at 
such a distance apart that the air is 
just able to withstand for an indefinite 
time. say. 10.000 volts maintained by a 
transformer, and then a strip of glass 
introduced between them, the insula- 
tion will break down, although the 
glass has greater dielectric strength 
than air. The explanation is quite 
simple : the fall of volts per centimeter 
of air is the highest the air can with- 
stand : as glass has a higher specific 
capacity the potential gradient in the 
"lass i> less steep than in the air, and 
the consequent increased steepness in 
the air punctures the latter. This ex- 
periment shows the necessity of hav- 
ing the insulation of uniform composi- 
tion. This does not exclude "graded" 
cables ; that is, those in which the small 
area in contact with the conductor is 
made dielectrically stronger than the 
peripheral areas. 


This method is a modification of one 
suggested in The Electrical Ace 
April, 1907. and is put forward tenta- 
tively in the hope that it will either re- 
ceive approval or bring forth sugges- 
tion s of improvement. 

The variety of opinion as to the 
proper thickness of insulation to be 
used shows that this' matter is not 
treated in a scientific way. The 
method of calculating the proper 
thickness, given below, while probably 
susceptible of improvement, is certain- 
ly much better than the haphazard 
guessing often employed. 

THE thickness of insulation. 

The thickness of insulation to be 
placed on a wire is governed by three 

1. Errors in size of wire, eccentric 
situation of wire in the insulation, and 
similar irregularities. 

2. Insulation not to be strained by 
application of test voltage. 

3. Insulation to be thick enough 
to have mechanical strength. 

error thickness. 

The thickness of insulation required 
to make up for errors and irregulari- 
ties of manufacture may be termed 
the "Error Thickness." This quantity 
depends both on the size of wire and 
on the thickness of insulation. While 
no exact expression of error thickness 
is possible, experience has shown it to 
be proportional to the square root of 
the wire diameter and proportional to 
the thickness of insulation. For rub- 
ber insulation the following empirical 
formula represents good practice : 

Error Thickness (inch)=o.oit-f- 


where D=diameter of wire, inches 

t=thickness of insulation, inches. 
Thus, for a Xo. 0000 B. & S. with 
0.28 in. of insulation. 

E=( 0.0 1 X0.28 ) -)- ( o.o~\/o.46) 


The formula may be more conveni- 
ently written : 

E=o.oit+o.098\/ r 
where r is the radius of the wire. 

dielectric stress. 

When a high potential is established 
across the insulation of a cable, the 
insulation is subjected to a strain 
which depends upon the degree of 
concentration of electric force. When 
this concentration reaches a certain 
value, the insulation will no longer 
be able to stand the strain and will 
break down. It will not necessarily 
be punctured, but will be disintegrated 
only where the concentration of elec- 
trical force has been excessive. For 
purpose of analysis, it is usual to 
represent the intensity of electric force 
by the density of imaginary lines of 
force stretching radially from wire to 

Let F=electric force or dielectric 
Y=Test potential, kilovolts 
t=thiekness of insulation, inches, 
over error thickness. 

then F=^-where the electric force is 
The electric force around a cylindric- 

al wire, however, is not uniform, the 
lines extending radially from the wire 
to the outside of the insulation. The 
density of the force lines is therefore 
greater at the surface of the wire than 
at the outside of the insulation. This 
explains the well-known fact that 
small wires insulated for high poten- 
tials often show a disintegration of the 
inner layers of insulation without any 
visible defect on the outside. In this 

p— .434V 

r log (t + r) 


where r is the radius of the wire, 
inches, and the logarithm is to the 
base ten. 

This gives: V=2.$026 F.r.log. 4 * 1- 

This is not strictly true for stranded 
cables, the dielectric stress being from 
1.23 to 1.46 times the value given by 
the above formula. 

The smaller value holds for thick 
insulation and the latter for very thin 

(The exact formula for stranded 
cables, according to Professor Levi- 
Civita, is given by E. Jon a in the 
Transactions of the International 
Electrical Congress at St. Louis, 
1904. ) 

Owing to the individual wires of 
a mutiplex cable being pressed to- 
gether in assembling, the voltage test 
between cables should be slightly less 
than double that on a single cable. 
Ten per cent, is taken as a conserva- 
tive amount for the loss due to this. 

electrical thickness. 

The error thickness being known, 
the electrical thickness, or thickness 
required to withstand the electrostatic 
stress, may be calculated with the aid 
of the logarithmic formula, above. 

The thickness of insulation adopted 
for potential differences up to about 
IOOO volts is determined solely by 
mechanical considerations, the dielec- 
tric stress not being concerned. 


The error thickness and electrical 
thickness of insulation are often in- 
sufficient for mechanical reasons. Fig. 
1 shows the minimum thickness o\ in- 
sulation which is permitted by me- 
chanical considerations. I'nlike the 
electrical thickness, which is added to 
the error thickness, the mechanical 




June, 190& 

thickness is a total figure which in- 
cludes everything. This graph, while 
based on average practice, may not 
meet the requirements of some engi- 
neers, and should, therefore, be care- 
fully examined before it is used. 


The insulation resistance of a cable 
is derivable from the following 
formula : 

M=4000 log 

■377+- 2 3 

M=58xio- 7 xSxlog T y r 


M=megohms per mile ; 
S=specific resistance in megohms 
per inch cube ; 

T=thickness of insulation inches. 
r=radius of wire, inches ; 
logarthim is to base ten. 

This formula is sometimes written 



K=58xio- 7 xS. 

The value of K varies from 870 to 
23,200 for S=i50 and S=4COO, re- 
spectively. The use of K instead of 
S has the advantage of brevity and is 
endorsed by the manufacturers. 

Calculating insulation resistance, 
the total thickness of insulation should 
be used. 


It is desired to find the thickness of 
insulation for a cable to be tested for 
15 kilovolts (using a stress of 96 kilo- 
volts per inch), the size being No. 
4-0 B. & S. stranded. 

Fifteen kilovolts, with stranded ca- 
ble, is equivalent to 15x1.3=19.5 
kilovolts with solid wire, say. 
Then using the formula 

Hog r=log (t+r) 



Hog -23=log (t+r) 


.383+1.361=^ (t+r) 

i.744=log (t+r) 



t=-556— -23=.326 

Error thickness, E=.oit+.c>98\/ J 
=.00326+. 0475 
E=.05i approx. 

Total thickness, T=.05i+-326 


The megohms per mile, assuming 
K=4OO0, are : 


= 1684 

In the above case the thickness is 
well above the amount required for 
mechanical strength, which would 
be about 6 / C4 inch. If the thickness 
had worked out to an amount less 
than is required for mechanical 
strength, the proper thickness would 
have to have been taken from Fig. 2. 

In such cases the error thickness 
has to be calculated and subtracted 

ages from 5000 up, and is very largely 
used for lower voltages. 

Owing to the hygroscopic qualities 
of paper insulation, it should not be 
used where the cable is exposed to the 
direct action of water, as. for ex- 
ample, in submarine work or in badly 
drained splicing chambers. For this 
service, rubber or varnished cambric 
insulation is to be preferred, as, in the 
event of a burn-out, the insulation 
will not be spoiled, except at the actu- 
al point of trouble. 

Dr. Jona, Int. Elec. Congress, 1904, 
says that paper subjected to dielectric 
strain for an hour, with progressively 


2 J 4 J .6 7 .9 /.t J/ // /J /4 16 U // /# /J 10 
Drtflre/tr 0/ /f/re, //tc/tes. 



from T in order to obtain t, for which 
the test voltage is calculated. 


127 kilovolts per in., conservative 

testing stress. 
56 kilovolts per in., conservative 

working stress. 
400 kilovolts per in., breakdown 

stress (approxr.) 



Paper ribbon is wound spirally 
around the conductor in numerous 
layers, until the desired thickness is 
obtained. The cable is then immersed 
in a bath of oily insulating compound, 
until saturated. The whole is then 
enclosed in a lead sheath, which not 
only serves to retain the compound, 
but also to exclude moisture. 

This type of cable is cheaper than 
varnished cambric or .good quality 
rubber, and is better able to stand 
high voltages than rubber insulation. 
It is almost universally used for volt- 

increasing voltage, will stand from 
eight to ten kilovolts per millimeter. 
These numbers represent good com- 
mercial averages, but it is not unusual 
to find paper with 20 or 30 per cent, 
greater dielectric strength. 


As the result of some fifteen years 
of experience with underground ca- 
bles, the following table, giving thick- 
ness of insulation and lead sheath for 
various sizes of conductors and work- 
ing pressures, is submitted as repre- 
senting conservative practice : 


Size of Conductors 

ness of 

Thickness op 

Single i Three 
Cond. Cond. 

No. 6 to No. 2 B. & S. . . 
No 1 to No. 00 

A in. 
A in. 
A in. 

A in. 


A in. : A in. 

A in. A in. 

No. 000 to 300,000 en... 
400,000 to 750,000 cm. . 
800,000 to 1,000,000 cm. 
1 ,250,000 to 2,000,000 en 

A in. A in. 
A jn. ; 

A in 

June, 1908 



For each iooo volts increase of 
pressure above 3000 add 1 / 32 -'m. in- 
sulation to the wafl until 11,000 volts 
is reached, and after that add x / 64 -in. 
for each 1000 volts. For example, 
the insulation required on a No. O 
B. & S. 25,000-volt cable would be 
20-32-in. or 5/^-in. If 35 per cent, 
para rubber compound or varnished 
cambric is used for insulation, the 
above empirical rule may be changed 
to read: for each iooo volts increase 
above 3000, add V 64 -in. insulation to 
the thickness of wall until 25,000 volts 
is reached. For the insulation of low- 
potential cables, V 32 -in. paper should 
be used on all sizes up to 1,000,000 
c. m., and from 1,250,000 to 2,000,000 
c. m., 5 /;, 2 -in. should be used. 

From a purely electrical point of 
view, one-half of this insulation would 
be ample to withstand 650 volts work- 
ing pressure, but the mechanical ef- 
fects of reeling and unreeling the ca- 
ble and pulling it into ducts and bend- 
ing around the manholes, are to prac- 
tically destroy the insulating quali- 
ties of the layer of paper next the 
lead, so that we really start in with 
a cable having approximately 1 / 32 in. 
of its insulation destroyed before it 
is put into commission ; this mechan- 
ical destruction of insulation is espe- 
cially marked in cold weather, as the 
oils used with the paper tend to con- 
geal when subjected to a temperament 
below 32 degrees F. The cable manu- 
facturers have met this difficulty by 
using more fluid oil, with the result 
that the insulation resistance of the 
cable may not be more than 50 meg- 
ohms at 60 degrees F., but by the use 
of this very soft insulation they have 
produced a cable giving a very low 
insulation, but a high puncture test, 
and at the same time have met, to a 
great extent, the difficulty of handling 
paper cable in cold weather. It is 
always advisable, however, if a cable 
is to be used in a temperature below 
32 degrees F., to keep it in a warm 
place, such as a boiler-room, for at 
least 12 hours before drawing it in. 

The cable may then be used in the 
coldest weather, as it gives up its heat 
very slowly." 

H. G. Stott, Oct., 1906. 


Temp. Deg. Cent. 

Factor for 
High Grade Paper 




5 20 






















] .57 


1 .36 



15.5 (60° F.) 







Time of Electrification, 

Relative Insulation 

Resistance, Referred to 

Value after I Minute 


o> . 



1 00 


1 09 










1 .31 


1 33 


1 .35 

This test represents average results, hut must not 
be taken as correct for any particular cable. 


Prepared cotton fabric is coated on 
both sides with multiple films of in- 
sulating varnish. The coated cloth 
is cut into strips and wound spirally 
on the copper core, with films of non- 
drving viscous adhesive compound 
between the layers. A separator is 
sometimes applied between the copper 
core and the taping, in order to pre- 
vent any possible action of the var- 
nished films on the copper. 

This insulation, unlike paper, does 
not absorb moisture and may be used 



New York Edison 

New York Metropolitan 

I. R. T. Subway (New York).. 
Manhattan Ry. (New York)... 
New York Central R.R 

Thickness of Insulation in Thousandths of 
an Inch 

Buffalo — Niagara Line. 

Chicago Edison. 


St. Paul 

St. Paul 



1 1 ,000 
1 1 .000 
1 1 ,000 

1 1 ,000 










Per IOOO Volts 
and Ground 



and Ground 







for indoor work without a lead sheath. 
It is suitable for high-tension cables, 
especially where a lead sheath cannot 
be used, as, for example, when sub- 
jected to vibration. In such cases it 
is usual to protect the insulation by a 
spiral galvanized steel tape. 

Cambric insulation is considerably 
more flexible than paper, it being pos- 
sible to bend cables to a radius of six 
times their diameter, without injury. 
Unlike rubber-insulated cables, the 
insulation remains concentric with the 

Other advantages of varnished 
cambric are that splices are simple, 
and that mineral oils have no effect 
on it. 


Rubber insulation, so-called, is a 
compound of various substances in 
which rubber seldom predominates. 
It is, therefore, not surprising to find 
the properties of rubber compounds 
varying between very wide limits ac- 
cording to the nature of the substances 
of which they are composed, and ac- 
cording to the process of compound- 

The qualities which a rubber com- 
pound should possess, in order to ful- 
fil all requirements as cable insula- 
tion, are as follow- : 

(i) High dielectric strength. 

(2) High mechanical strength. 

(3) Fair elasticity. 

(4) Fair specific resistance. 

(5) Permanence or long life. 

The first four qualities are not diffi- 
cult to obtain, and it is easy to test 
a compound for their presence. The 
fifth quality, permanence, depends on 


Temp. Deg. Cent. 

Factor for 
Varnished Cambric 














5.00 * 






2 7<i 











15.5 (60° F.) 




♦This varnish eamb ic < able has a double steel tape without lead and is used in pipes on elevated 

two conditions: The first of these is 
chemical equilibrium, i. e., the rubber 
and substances associated with it must 
have no affinity for one another, for 
the conductor, for air or for moisture. 
The second condition i< that the com- 
pound shall contain no sub>tance 
tending to change its physical state. 
as, for example, a volatile or crystal- 
lizable substance. 

Within wide limits compounds oi 
various compositions can be made 



June, 1908 


Working Voltage 

1000 Volts 
or Less 

3000 Volts 
or Less 

5000 Volts 
or Less 

7000 Volts 
or Less 

Test Volts 

6 B. & S 







H M. C. M . 

.3 M. C M 

4 M. C. M 

1 - M. C. M 

3 4 M. C. M 

1 M. C. M 

1 M M. C. M 

1 ! . M. C. M 

2 " M. C. M 








Test Volts 







Test Volts 

Test Volts 








15,000 Volts 
or Less 

Test Volts 


These are the figures recommended by the G. E. Co. 


Working Voltage 


or Less 

or Less 

5000 7000 
or Less or L 

or Less 

or I. 

Test Voltage 



12.500 17 5i)i 


33 000 



Inches Inches 



A- A 
A - A 

A- A 

ft - H 

a -a 


a -a 

The first column in each group is the thickness of insulation on each conductor, and the second is the 
thickness over all. 

These figures are recommended by the G. E. Co. 










































. •>. 

/fb/e = j?6 %jc#r a(y/&9 /»Ar 



r t 






JV ■#) SO 60 7a SO SO M 

Temperaft/re, degrees /5/ve>*6e//. 


M W 




balanced and, therefore, permanent, 
provided that conditions inconsistent 
with the condition of balance are not 

specified. There are no known tests 
which will infallibly distinguish be- 
tween a balanced and an unbalanced 

compound. A short discussion of the 
tests and restrictions which have been 
suggested for this purpose is given 


Rubber is a gum extracted from a 
tree which grows in the tropical coun- 
tries of Africa and South America. 
The quality of this gum varies in 
many ways, but the characteristic 
which most affects its commercial 
value is the amount of resinous ex- 
tract which it contains. The amount 
of extract is usually estimated by di- 
gesting the gum in acetone for several 
hours, and thereby dissolving out the 
extract. The proportion of acetone 
extract in different grades of gum 
varies from less than i per cent, to 
over 20 per cent., the grades having 
the smaller proportion of extract be- 
ing generally from South America. 

The best grade of South American 
rubber is known as fine Para, and is 
the most desirable kind to use in in- 
sulating compounds. While it is usual 
to specify that compounds shall con- 
tain only the finest dty Para rubber, 
fliers is no practical way to frscvrtain 
whether the rubber did actually come 
from Para. Furthermore, it is of no 
practical import whence the rubber is 
from, provided that the percentage of 
extract does not exceed, say. 3 per 
cent. A greater percentage of extract 
indicates a cheap grade of rubber, 
which it is difficult to manufacture 
into a balanced compound. 


Rubber gum. in its native state, is 
of little use for insulating purposes, 
owing to its property of absorbing 
water and oxidizing. When mixed 
with sulphur and heated to a tempera- 
ture of from 248 to 302 degrees fahr.. 
a combination takes place which ren- 
ders the rubber more stable and at 
the same time increases its mechan- 
ical and electrical strength. This 
process is known as vulcanization. 


It has been found by experience 
that 60 to 70 per cent, of adulterant 
may be added to rubber gum without 
destroying its useful qualities after 
vulcanization. Above this percentage, 
the qualities of the rubber cease to 
predominate, and the compound par- 
takes markedly of the characteristics 
of the adulterant. It is for this reason 
that 30 per cent, pure rubber is gen- 
erally adopted as the standard propor- 
tion, and that 40 per cent, pure rubber 
is required for shipboard work in the 
navy, the larger proportion being 
adopted as a special precaution on ac- 
count of the necessity of absolute re- 

June, 1908 




A good 30 per cent. Para com- 
pound, properly vulcanized, should 
show a tensile strength of at least 
800 pounds per square inch. This 
figure is agreed to by practically 
every manufacturer of rubber com- 
pound in the United States, but the 
proportion of compounds which actu- 
ally show this tensile strength is small. 

A sample should be cut so that the 
ends gripped shall be considerably 
larger than the center, where the 
break should occur. The sample 
should be bent slightly, in every di- 
rection, before testing, in order to 
magnify and reveal any surface in- 
cisions which might reduce the total 


When stretched three times its orig- 
inal length, a sample should show a 
set not greater than 18^4 per cent, 
after a stated time has elapsed. Al- 
though the time is a matter of con- 
troversy, this percentage set is agreed 
to by all the leading manufacturers. 

Nevertheless, it is well known that 
certain excellent compounds entirely 
fail to meet the regular stretch tests. 
It is, perhaps, better to lose the use of 
this class of compounds and take ad- 
vantage of the selective action of the 
stretch test; and if this is done, it 
should be specified that the test may 
be performed by the purchaser at any 
temperature between 50 and 100 de- 
grees fahr. It should also be speci- 
fied that the sample tested shall not 
have been submitted to any previous 
stretching, because a sample with a 
permanent set will not show much ad- 
ditional set when further stretched. 
Stretching should be steady and re- 
lease instantaneous. 


The specific resistance of insulation 
sold as 30 per cent. Para compound 
varies between the enormouslv wide 
limits of 150 millions of megohms per 
inch, cube and 4000 millions of meg- 
ohms per inch cube. 

From the standpoint of leakage, a 
mere fraction of the smaller value 
would be sufficient. It is, therefore, 
only as a test of quality that high meg- 
ohms may be demanded, and the 
value of such test is open to doubt. 

A minimum of 750 millions of meg- 
ohms per inch cube is conservative, 
and there is nothing to be gained by 
specifying over 1200 millions of meg- 
ohms per inch cube. 


The rate of change of resistance 
with regard to temperature should not 
exceed 2.6 per cent, per degree Fah- 
renheit. This is in agreement with 

the tables used by the most reputable 
manufacturers. The object of speci- 
fying this quantity is twofold : First, 
to prevent the manufacturer using any 
temperature correction factor which 
will give a figure which complies with 
the specifications; second, as a meas- 
ure of quality of the compound as 
pointed out by H. G. Stott, Proc. Am. 
Inst. Elect. Eng., 1906. 

The writer's experience very strong- 
ly confirms Mr. Stott's opinion of the 
value of this test. 


If extensions and retractions are 
plotted on a base of load, a complete 
"hysteresis" loop is obtained, as shown 
in Fig. 3. The area of this loop 
should generally be small in good com- 
pound ; there are. however, exceptions 
to this rule. 


Sulphur in rubber compound may 
be in three conditions : 

( 1 ) free ; 

(2) combined with rubber ; 

(3) in barium sulphate. 

Poor quality rubber requires a great 
deal of sulphur to vulcanize it. and 
is. therefore, often revealed by the 
large amount of combined sulphur. 

Excess of free sulphur, say, over 
1 per cent., usually indicates an un- 
stable compound, as the sulphur is 

liable to combine with the copper or 
tin coating over the copper. 


Brand op Ri'bber 



Upper Congo 


Sierra Leone . 

Resin in 



Per Cent. 

4 04 



C. O. Weber, Chemistry of India 


C. O. Weber (Chemistry of India 
Rubber. London. 1903) defines the 

SIr«S3 . Lbi per Sq. in. of Original A rea . 

coefficient of vulcanization as the per- 
centage ratio of amount to rubber and 
sulphur of vulcanization : 

30% Rubber Compound. Megohms Per Mile. 60 Deo. Fahr. One Minute Electrification 




5 64 




5 32 



1 ,000,000 cir. mils 















































1 ,380 




























900,000 '• 




700,000 " 


600,000 " 


500 000 " ' 





300.000 " 




4 stranded 


. 905 



3 " 


2/0 " 




1 solid 





3 " 






6 " 


8 '• 



9 • 


10 •' 

1 .620 

12 " 




30% Rubber Compound. 

Voltage Test for 5 Minutes. 


For 30-Minute Test Take 80% of The^e 


1. 000.000 to 550,000. 
500,000 to 250,000. . . 
4/0 to 1 

2 to 7 

8 to 14 

Thickness of Insulation 


4 64 



5 64 










4 32 ' 5/32 

6 32 7 32 

9 32 

6.000 10000 14.000 18,000 22,000 26.000 

8,000 12,000 16,000 20,000 24,000| 28,000 

10.000, 14,000 18,000 22,000 26.000 30,000 

12,000 16,000 20,000 24.000 28.0001 32,000 

13,000 17,000 21, 000 25.000) 

10 32 


32, OOO 

]\". S. Clark, Am. Inst. Elect. Eng.. 1906. 



June, 1908 


Coefficient of 

Extension — Inchus 


Load, Lbs. 



considerable amount of free sulphur. 
(3) Stickiness and darkening in color. 
Rubber containing mineral oils, 
large quantities of recovered rubber, 
or large proportions of sulphide sub- 

The extension follows a straight-line law only with 
considerable loads. 


(1) Loss of strength or cohesion. 
Rubber with a low coefficient of 

vulcanization is liable to develop this 
defect, particularly if the time for vul- 
canization has been short. 

(2) Hardening with brittleness. 
Rubber may contain white substi- 
tutes (chlorosulphides), but more 
commonly is due to the presence of a 

Temp. Cent. 







Slightly sticky. 

Sticky, but slightly elastic. 

Surface melts and rubber 

Gradually melts. 

Can be mixed up and ther- 
mometer easily pushed 
into the mass. 

Appearance of decomposi- 
tion and boiling. 

Gas evolved, which burns 
with a luminous flame. 

Insulation Resistance and Puncture Tests (30% Para Compound) 


B. & S. Gauge 

Voltage Test 
Wall for 1 Minute 

Insulation Resistance 

Xo. 14 to 8. .""!". 

A in. 1,000 

,', ■■ 1,000 
A " 1.000 . 

A ' 1,000 

A " 1 000 

1,000 megohms per mile 
1 ,000 

'■ 6 to 2 

" 1 to 4/0 

250.000 to 500,000 cir. mils 

550,000 to 1 ,000.000 " 



B \ S. Gauge 


Voltage Test 
for 1 Minute 

Insulation Resistance 

Xo. 14 to 8 

A in- 

A " 


3.000 megohms per n.i'e 


" 1 to 4/0 


250,000 to 500.000 cir. mils 


550,000 to 1,000,000 " 






B. & S. Gauge 


Voltage Test 
for 1 Minute. 

Insulation Resistance 

No. 4 to 4/0 

250.00C to 500.000 cir. mils 
550.000 to 1,000.000 




2,500 megohms per mile 





B t V-JS. Gauge 


Voltage Test 
for 1 Minute 

Insulation Resistance 

No. 4 to 4/0 

A in. 
A " 
A " • 


4.000 megohms per mile 

250.000 to 500.000 cir. mils. . 


550.000 to 1 ,000,000 " 


B. &"S. Gauge 


Voltage Test 
for 1 Minute 

Insulation Resistance 

No. 4 to 4/0 

250,000 to 500.000 cir. mils 
550,000 to 1,000,000 

B. &^S. Gauge 

20,000 5.000 megohms per mile 

20.000 4.000 

20,000 2,500 


Voltage Test 
for 1 Minute 

No. 4 to 4/0 

250,000 to 500.000 cir. mils 
550.000 to 1,000,000 , 

Insulation Resistance 

20,000 6.000 megohms per mile 

20,000 5.000 

20,000 3.000 

The liquid obtained on heating be- 
comes viscid on cooling, but it does 
not again solidify. 


Temperature in Deg. Fahr. 

Loss of Tenacity. 
Per Cent. 















A. Schwartz, Journal Inst. E. E., 


Rubber overworked in the masti- 
cator oxidizes very rapidly, yielding 
a much greater amount of extract than 
before mastication. 

C. O. Weber, Journal of Society of 
Chemical Industry, 1903, p. 875 and 
p. 103. 


The action of light on rubber, 
whether vulcanized or unvulcanized. is 
an oxidizing action, but the oxidation 
is faster the lower the degree of vul- 

C. O. Weber, Journal of Society of 
Chemical Industry, 1903, p. 875. 


The deterioration of Congo rubber 
is due to the presence of albuminous 
substances primarily. Coagulated al- 
bumin is not removed by washing, 
causing finished goods to be more or 
less brittle, according to the amount of 
albumin present. 

C. O. Weber, Journal of Society of 
Chemical Industry, 1902, p. 712. 


Certain varieties of rubber do not 
become properly vulcanized when 
treated with sulphur only, but do so 
readily if a considerable proportion of 
litharge is present during the process. 
The effect of litharge, however, is to 
make the rubber brittle. 

C. O. Weber, Journal of Society of 
Chemical Industry, 1903, p. 103. ■ 


Insulation Resistance and Puncture Tests 

(30% Para Compound) 

B. &S. 


Test for 


1 Minute 



A in. 


3,000 Megohms 


A - 







A " 




A " 




A " 




A " 




A " 


2 500 


A " 




A " 




A " 



A " 




A " 




A " 




A " 



/. Lankan, Am. Inst. Elect. E 

J. Langan, Am. 

Inst. Elect. Eng. 



Compensators for Measuring Line Drop 


FOR large lighting systems it is 
often of great importance to keep 
the pressure constant at the va- 
rious distributing centers. In order 
to do this it becomes necessary to 
know the distributing-pressure in the 
power-station so that the proper regu- 
lations can be made. 

In earlier days it was customary to 
run independent pressure wires from 
each distributing center to the central- 
station. With long-distance lines, 
however, the cost of the pressure wires 
becomes excessive, and some other 
means had to be resorted to. For this 
reason the compensator is now-a-days 
used altogether. This apparatus serves 
to modify the reading of the station 
voltmeter without the use of pressure 
wires so that the reading corresponds 
to the pressure at the point of distribu- 
tion. The compensator must be so 
connected and adjusted so as to allow 
for the resistance and reactance of the 
line for which it is used. It is neces- 
sary to provide such an electromotive 
force at the power-station so that the 
voltmeter under the influence of the 
compensator will give the same read- 
ing as if connected directly at the dis- 
tributing point, independent of the 
current and the power factor. An elec- 
tromotive force component must be 
obtained which is in proportion to and 
in phase with a line drop and also 
has the proper relations to the electro- 
motive force in the power-station. 

Compensators used for single phase 
or balanced two or three-phase sys- 
tems are connected to one-phase only 
by means of a single current trans- 
former. For unbalanced systems one 
compensator for each phase is used to 
advantage. For high pressures the 
use of potential transformers also be- 
comes necessarv. 


V=Voltmeter. 9 

D=Center of distribution. 

R=Ohmic resistance of line. 

X=Reactance of line. 

R 1 =rOhmic resistance of compensa- 

X^Reactance coil of compensator. 

E=Generator pressure. 

I=:Current in the line. 

I x :=Secondary current in current 

A is the generator in the central- 
station generating the pressure E. 

X x in the compensator, it causes the 
same drop in the electromotive force 
of the voltmeter circuit as the line 
wires cause in the electromotive force 
of the load. 

In the following is shown a method 
for determining the values of Rj and 
X 1 = 

R=Ohmic resistance of line 

X=Reactance of line 

I=Line current 

p=Ratio of potential transformer 

fig. 2. 

This pressure is reduced in the ratio 
E : E x by the potential transformer P. 
In the same manner the current is 
transformed in the ratio I : I x by the 
current transformer C. In parallel 
with the secondary circuit of this 
transformer is connected an ohmic re- 
sistance R x in series with an induction 
coil of Xi ohms reactance. The ohmic 
resistance of the line is R and the re- 
actance X. Then, by properly adjust- 
ing the resistance R x and the reactance 

c=Ratio of current transformer, 



= — and C= 



R 2 +X 2 =Zz 




Impedance of the 

\/R 1 2 H-X 1 -.I 1 =\"oltage drop in 
the voltmeter circuit. 

Z.I E 


fig. 1. 

In Fig. 1 is shown a diagram for a 
compensator as used on a high-tension 
single-phase circuit. 


P=Potential transformer. 

C=Current transformer. 

\/R 2 +X 2 . 1 = Voltage drop in the 

V / Rv 2 +X 1 2 =Z 1 =Impedance of the 
voltmeter circuit. 
















From these last two equations the 
values of R t and X t can be calculated. 
E x and Z^Ij are in phase with E and 




June, 1908 

Z . I and the reading of the voltmeter 
V is proportional to the pressure at 
the distributing point for every cur- 
rent and power factor in the line. 

A great advantage of this system is 
that all apparatus is connected in the 
secondary circuits and that all danger 
from coming in contact with high- 
tension currents is thus eliminated. 

If it becomes desirable to measure 
the pressure at more than one distrib- 
uting center, this can easily be done 
by using one common potential trans- 
former with voltmeter and a throw- 
over switch to connect the compensa- 
tors for each outgoing line. One cur- 
rent transformer and one compensator 
is necessary for each line. Fig. 2 
shows a diagram for this arrangement 
which explains itself. 

An example will be given showing 
the method of calculating the com- 
pensator adjustment for a certain 

The system is three-phase with bal- 
anced load. The distance between 
the generating-station and the point 
of distribution is 20 miles. The line 
consists of three copper wires of No. 
4 B. & S. gauge strung 48 in. apart. 
The prfssure at the generating-station 
is 30,000 volts and the line current 75 
amperes. Frequency 60 cycles. 

The cross section of No. 4 wire is 
approximately 41,750 cm. and the 
radius of the wire approximately 0.1 
in. Specific resistance of copper is 

Then the resistance of one phase of 
the line is : 

R= =27.2 ohms. 


The inductance for 20 miles of one 
phase of the line is : 

48 \ 
-740 log.— J] 
0.1 / 

1^=20(80.5+740 log.— )io — c =.04i6 


The reactance X of the line is : 

X=2XtX6oXo4i6=i5.8 ohms. 

If the ratio of the transformers is 

p=— =300 


c=— =15 


c 15 
R x = — R= X 27. 2 — 1. 36 ohms. 

p 300 

c 15 

Xj= — X = X 15-8=0.79 ohms. 

I> 300 


Question. — We have a shunt 110- 
volt motor running a blower. Acci- 
dentally, when working at the rheo- 
stat while the motor teas running, I 
broke the field-circuit at the lug. 
When we started up again I found the 
armature injured, and on testing lo- 
cated a punctured coil. As the arma- 
ture was all right before I touched 
the iield, I cannot make out how my 
breaking a field circuit could have 
hurt the armature, even if my boss 
says I must have, in some way. been 
the cause of it. 

Answer. — Never open a field-circuit 
suddenly. If you are familiar at all 
with dynamos, you must know that 
current is generated by a wire cutting 
lines of force. The field-coils of a 
motor build up an immense field of 
these lines of force. When the cur- 
rent was first turned into the fiekl- 
coils, it took possibly 1 / \ -sec. for the 
field to build to the full value from 
the source of no V. If by breaking 
the field-circuit you tended to destroy 
the field of force in the V100 P art OI a 
circuit, then the field of force would 
collapse ten times as fast as they built 
up. As the rate of cutting determines 
the voltage the armature would have, 
each conductor suddenly generates 
110x10=1100 volts. Depending on 
this rate of collapse the momentary 
voltage might run into several thou- 
sands. An action of this kind took 
place in your case, hence a puncture. 

Question. — What is the reason for 
U'cstinghouse rotary converters liav- 
ing copper rings on the pole tips? 

Answer. — To prevent hunting and 
to help synchronous running. Hunt- 
ing is due to surges on the line, and 
may be up or down. These copper 

rings, known as dampers, are ordi- 
narily in a light, even, magnetic field, 
and are practically inert. A surge 
back into a rotary will change this 
condition, and either suddenly build 
up or tear down the field. This re- 
sults in a big moving field cutting the 
heavy rings, causing a correspond- 
ingly heavy current to be induced in 
them. This action reacts by opposing 
the cause of it, by choking down an 
upward rush, or holding up a falling 
(me. It thereby tends to keep condi- 
tions normal. 

Question. — We need more trans- 
formers for our outside lines, but on 
account of the hard times cannot af- 
ford to spend much money on them. 
We have considered buying trans- 
formers which will be heavily over- 
loaded during the peak period, but 
will have a fair load at other times. 
This will be cheap and the core loss 
will be reduced. Do you not think 
this would be a good plan/ 

A n s w e r. — M o s t manufacturers 
make transformers to give the most 
efficient output at 75% to 100% load. 
Running at light loads or heavy over- 
loads would not be economical, and 
would more than offset running them 
part of the time at highest efficiency, 
with a small core loss during light 
load. But the most objectionable fea- 
ture would be the danger of injury 
due to excessive overload. A big re- 
pair bill, or a new transformer, would 
take up any possible savings for a 
long while. Temperance is the great 
lesson of life, and applies to buying 
machinery as well as to human af- 
fairs. Don't be penm wise and pound 
foolish by purchasing something your 
judgment does not fully sanction. 

Question. — / have four electric bells 

in series which I cannot get to work. 
I followed a suggestion in an instruc- 
tion book, and turned them all into 
single-stroke bells, except one. When 
the button is pushed there is a bit of a 
tinkle sometimes, but no ring. What 
is probably the cause of my trouble? 
The circuit tests are 0. K. 

Answer. — The trouble is due to 
faulty adjustment of the screws which 
regulate the stroke of each bell. All 
bells must be adjusted for the same 
movement as the master bell. 

Lamp Testing 

Care should be taken that good 
contact is secured between all lamp 
bases and their holders. It is also im- 
portant to connect the voltmeter leads 
as near as possible to the lamp ter- 
minals. Before closing the standard 
lamp circuit sufficient resistance 
should be introduced to avoid all pos- 
sibility of burning the lamp at a volt- 
age higher than its rated voltage. 
After this circuit is closed, the volt- 
age should be gradually increased to 
the standard voltage and the current 
consumption should then be observed. 
If the current flowing is then more or 
less than that specified in the stand- 
ardization certificate, either the meters 
are incorrect or the lamp has changed. 
Reference to the other standardized 
lamps of the series will indicate which 
is the case. If these agree well among 
themselves and the discrepancy still 
exists, one instrument may be as- 
sumed correct and the other corrected 
so* as to bring the wattage up to 
standard. All lamps to be tested 
should be mounted with the axes of 
their filaments perpendicular to the 
photometer bar. 

General Electric Report 


E. W. Rice, Jr., Vice-President of 
the General Electric Co., reporting to 
President Coffin on the engineering 
work of the past year, says : 

"During the first part of the last year 
our engineers were fully occupied in 
supervising the technical details of our 
great ly expanded business. Upon the 
decline in business which followed 
they have had more time to devote to 
improvements and economies in design 
of our apparatus. More attention has 
also been given to the design of special 
apparatus intended to meet novel con- 
ditions and to the extension of our 
business along profitable lines. 

"The apparatus designed by our en- 
gineers for the long-distance transmis- 
sion of electricity has proved most re- 
liable, economical and satisfactory- in 

"There has been a continued in- 
crease in the capacity of electric gen- 
erators and transformers. 

"Our high-tension switching ap- 
paratus has been still further im- 
proved, and we have been favored 
with the most important orders for 
such installations. 

"The details of our steam turbine- 
generators have been improved, great 
economy and proved reliability are 
now assured, and the turbine-genera- 
tor is now standard for all new im- 
portant electrical installations where 
steam is utilized. We are now build- 
ing turbine units of a capacity of 14,- 
000 kw. ; the largest electrical generat- 
ing units ever produced. The Com- 
monwealth Edison Company, of 
Chicago, has now in operation in one 
station nine large turbines capable of 
generating a total of 103,500 kw. 

"Our engineers have devoted con- 
siderable attention to the design of a 
line of turbine-generators for use with 
exhaust steam. Such steam turbines 
are so much more efficient than steam 
engines when operated by low-pressure 
steam that they can be most usefully 
employed to supplement steam engines 
in existing installations. Their use 
will result in large increases in output 
without any increase in coal consump- 

"Our single phase alternating cur- 
rent railway equipments have been 
greatly improved during the past year. 

"Our new direct current railway 
motor, mentioned in my last report, 
has proved so satisfactory in practical 
operation that it is rapidly being 
adopted as the standard type. It 

marks an important advance in econo- 
my and durability. 

"We have extended the range of 
economical operation of direct current 
railway apparatus by designing it for 
use at 1200 volts, about double the 
existing standard, and have sold a 
number of such equipments to the 
Southern Pacific R. R. Co. 

"We have sold to the Great North- 
ern R. R. Co. four 100 ton three phase 
electric locomotives designed to handle 
all trains traversing the 2^2 miles of 
Cascade tunnel in Washington. This 
installation will be especially notable as 
the first instance of the substitution of 
electricity for steam on a mountain 
division of one of the Continental rail- 
ways. The traffic conditions are pecul- 
iarly difficult on account of the grades 
and tunnels. These electric locomo- 
tives, because of their increased speed 
and better control, will practically 
double the traffic capacity of the pres- 
ent steam locomotives. Electricity for 
their operation will be supplied from 
water power hitherto unused. 

"A gas-electric car, which fully 
meets the requirements of Steam Rail- 
road Companies for service on branch 
lines, has been perfected. The equip- 
ment consists of a gasoline engine 
driving an electric generator, which 
furnishes current to standard railway 
motors. The engine and generator are 
located in the forward end of an espe- 
cially designed car conveniently di- 
vided into passenger and baggage 
compartments, making a complete self- 
contained unit. 

"We have made many valuable im- 
provements in the design of machinery 
for electric reduction of metals and in 
apparatus for various industrial ap- 

"We have shipped several large mo- 
tors of special design of about 10,000- 
h.p. capacity each for driving rolling 
mills, and have received orders for ad- 
ditional equipments. 

"Important improvements in the de- 
sign of our lines of wiring devices, 
rheostats, circuit breakers, switches, 
instruments, and other small devices, 
have been made during the year. 

"Our new tungsten incandescent 
lamp, which gives more than double 
the illumination of the carbon filament 
for the same expenditure of power, has 
been further developed and has now 
become a standard commercial article. 

"Several novel types of arc lamps 
of greatly improved economy have also 

been perfected and sold in large quan- 


The General Electric Company's an- 
nual report upon sales says : 

"Among many important orders re- 
ceived during the year are : 

"Great Western Power Company, 
San Francisco, Cal. ; three water- 
wheel generators, 10,000 kw. each, to- 
gether with the necessary transform- 
ers and other electrical apparatus for 
transmitting current at 100,000 volts 
from its power house on the Feather 
River to Oakland, Cal., a distance of 
about 165 miles. 

"The Central Colorado Power Com- 
pany, Colorado Springs, Colo. ; four 
5000-kw. generators and other elec- 
trical apparatus for water power de- 
velopment at Glen wood Springs on 
the Grand River, the electrical energy 
to be transmitted throughout the cen- 
tral portion of the State for mining, 
general power, lighting and railway 

"The Detroit River Tunnel Com- 
pany, a subsidiary of the Michigan 
Central Railway Company; apparatus 
for equipment of the Detroit Tunnel 
under the St. Clair River. The con- 
tract includes several 1000-kw. motor- 
generator sets with accessories and 
six 100-ton locomotives, each equipped 
with four 250-h.p. motors. 

"The Great Northern Railway, for 
electrification of the Cascade Tunnel : 
water-wheel generators and 100-ton 
locomotives, each equipped with four 
250-h.p. alternating-current motors, 
giving a continuous output of iodo- 
Il.p. per locomotive. 

"The Southern Pacific Railroad 
Company, for electrification of its sub- 
urban lines in Oakland and Alameda. 
Cal. ; forty-four four-motor equip- 
ments with Sprague-General Electric 
control. The motors are 125 h.p. each. 

"The Hudson Tunnels Company : 
the turbine generators, rotary con- 
verters, motors and controlling appa- 
ratus for complete electrical equip- 
ment of its system of tunnels under 
the Hudson River connecting New 
Jersey and Manhattan. A portion of 
this system was put into successful op- 
eration on February 25, 1908, and reg- 
ular service is now maintained be- 
tween 19 Street (6th Avenue), New 
York City, and Hoboken, N. J. 

"The West Jersey & Sea Shore R.R. 
Co., a branch line of the Pennsylvania 
Railway running from Camden to At- 
lantic City, mentioned in last year"s re- 

J 35 

J 36 


June, 1908 

port. has maintained its record of satis- 
factory operation and orders for addi- 
tional equipment have been received 
during the past year to provide for the 
increased traffic. 

"The New York Central and Hud- 
son R. R.R. Co. is now operating in 
its New York City Terminal 35 elec- 
tric locomotives of our manufacture, 
each equipped with four 550-h.p. di- 
rect-current motors. Twelve addi- 
tional locomotives have recently been 
ordered, making a total of 47 locomo- 
tives purchased from us by this com- 

'The use of electrical apparatus for 
industrial purposes is extending rap- 
idly and large purchases of our appa- 
ratus have been made during the year 
for completely equipping mills with 
turbine and engine-driven generators 
for lighting and power, and with mo- 
tors of standard and special design for 
driving machinery of every descrip- 

"Orders for supplies, such as me- 
ters, transformers, arc lamps, wiring 
devices, electric heating devices, re- 
pair parts of electrical apparatus, etc., 

show an increase over last year. Our 
list of supplies comprises upward of 
50,000 items, separately catalogued 
and priced. In addition to the large 
stock of finished product carried at the 
several points of manufacture, we 
maintain 14 warehouses in various 
cities from which shipments to the 
value of over $6,000,000 were made 
during the year. 

"1200-Volt Direct Current System. 

"To meet the requirements of inter- 
urban railways where a potential high- 
er than 600 volts is desirable and the 
conditions are unfavorable to the 
adoption of the single-phase alterna- 
ting-current system, we have devel- 
oped a high-voltage direct-curr e nt 
railway system to operate at 1200 
volts. Two roads have been operating 
under this system for several months 
with entire success. Equipment for 
several additional roads of this char- 
acter is in process of installation. 
"The Curtis Steam Turbine. 

"The Curtis steam turbine contin- 
ues to give excellent service, and the 
confidence of users is evidenced by 
numerous additional orders for exist- 

ing installations. The total number of 
Curtis turbines shipped to date is 960, 
having a total capacity of 1,086,000 h.p. 
Orders were received during the year 
for turbines aggregating 380,000 h.p. 
We now have in process of manu- 
facture for the Commonwealth Edi- 
son Company of Chicago and the New 
York Edison Company a number of 
turbine generators of 14,000-kw. ca- 
pacity each, which will be the largest 
steam-driven electrical units ever pro- 
"Incandescent Lamps. 

"The consumption of carbon fila- 
ment lamps has steadily increased dur- 
ing the year, and with our enlarged 
capacity we are prepared to take care 
of the demand. In addition, we have 
received large orders for different 
types of high efficiency metal filament 
lamps, first consideration being given ■ 
to such sizes and types as will aid cen- 
tral lighting stations in providing for 
the requirements of their customers 
and the extension of their business." 

The actual condition of the Com- 
pany and its allied companies and 
their operations is shown in the fol- 
lowing balance sheets : 


Patents, Franchises and Good-will S 1 00 


Cash 1l'.l>50,720 92 

Stocks and Bonds $18,000,089 85 

Real Estate (other than factory Plants) 541.900 50 

Notes and Accounts Receivable 29,857,726 84 

Work in Progress 1,276,294 22 


5 % Gold Coupon Debentures of 1892 

3J% " of 1902 

of 1907 

Accrued Interest on Debentures 

Accounts Payable 

Unclaimed Dividends 

55.000 00 

2 047.000 00 

12,872,750 00 

108,791 67 

1,759,517 47 

1.469 86 

$16,844,529 00 

Merchandise Inventories: 

At Factories $18,339,652 06 

At General and Local Offices 2,422,678 59 

Consignments . ... . . r. .- 234,725 16 

S49.676.011 41 

20,997,055 81 

Factory Plants (including all lands, buildings and 

machinery) $12,900,000 00 

Copper Mining Interest 2,701 ,976 00 

70,673,007 22 Capital Stock Issued. 

15,601.976 00 Surplus 
$98,525,765 14 

65.167,400 00 

16,513,836 14 
$98,525,765 14 


Cost of Sales (including depreciation of Plants $3,745,989.06) $65,536,305 

Interest on Debentures 362,029 

Profit for the current year 6,586,653 




-72 484,988 06 

Dividends paid in Cash $5, 183,614 00 

Surplus at January 31, 1908, carried forward to next year 16,513,836 14 

$21,697,450 14 



Royalties, Dividends, Bond Interest, readjustment 

Stocks and Bonds account, and Sundry Profits. SI ,010,961 
Interest and Discount 487,079 

Profit on Sales of Stocks and Bonds 

.$70,977,168 46 



1,498.040 67 
9.778 93 

S72,484,988 06 

Surplus brought over from last year $15, 110,796 77 

Profit for the year ending January 31. 190S 6,586.653 37 

$21,697,450 14 


Property Accounts. _. S4.523.2S4 93 

Patents, Franchises and "Gobd-will . . ' " 3 00 

Current Assets 

Merchandise, Material and Supplies. , $2,560,100 01 

Work in Progress 114,389 20 

Notes- and Accounts Receivable 1,642,752 98 

Stocks and Bonds 7.426 70 

Cash 301,782 43 

Discounted Paper. . 


4,626,451 32 
246 48 

$9,149,985 73 

Capital Stocks $4,015,000 00 


Current Liabilities 

Genera! Electric Company 


As at January 31, 1907. $1,309,982 23 

Add profits for year $365,438 94 

Less dividends 160,000 00 

205, 43S 94 


535,000 00 
223,382 42 
2,860.935 66 

1,515,421 17 
246 48 

$9.149,9S5 73 

HENRY W. DARLING, Treasurer. 
EDWARD CLARK, General Auditpr. 

June, 1908 



New BreaKdown Rate for 
New Yorh 

As a result of the investigation con- 
ducted by Commissioner Milo R. 
Maltbie and according to the recom- 
mendations made bir him in a prelim- 
inary report to the Public Service 
Commission the New York Edison 
Company has modified its rates for 
"break-down service" so that the 
charge shall be based upon the max- 
imum demand of the consumer and 
not upon the full capacity of his in- 
stallation. This agreement is to last 
for one year, during which time it is 
also agreed the company will place 
recording devices upon every break- 
down connection and take the records 
of the current furnished under the 
supervision of the Commission, for the 
purpose of determining a fair charge 
for the service. 

Commissioner Maltbie finds that a 
satisfactory decision of the rate ques- 
tion cannot be made until such records 
are taken, and he recommends that the 
work be carried out and that the elec- 
trical engineer of the Commission be 
instructed to supervise it. The fact 
that no such records have been taken 
in the past makes it impossible to ar- 
rive at a just and adequate charge for 
this kind of service. 

The report points out that "break- 
down service" includes three kinds of 
service : 

No. i. — The service supplied by the 
•company to a private consumer having 
his own plant, when that plant breaks 
down and is unable to produce the 
current required by the consumer. 

No. 2. — Auxiliary service supplied 
to private consumers using current at 
nights, on Sundays and holidays, and 
during periods when a small amount 
of current is used as compared with 
the total installations. • 

No. 3. — Peak load service — addi- 
tional current furnished to private con- 
sumers at that hour of the day when 
the peak load or the maximum demand 
from all consumers occurs. 

Commissioner Maltbie says that the 
term "break-down service" does not 
include the supply of current for a 
segregated circuit in any building, and 
that where such segregated circuits 
exist, even though there may be a 
private electric plant tfpon the prem- 
ises, all of the electric light companies 
have stated that they are willing to 
supply current at the usual rates. 
. As to the first class of "break-down 
service," the report says it is probable, 
in view of the comparative in frequency 
of complete disability, "that the actual 
demands made upon the supply com- 
pany would be very few and would 
not involve great expense, either in 
the way of fixed charges upon the 

duplicate plant kept ready for use, or 
in the way of operating charges for 
units revolving slowly." 

In order to furnish such a service 
the supply company must provide 
merely the distribution system and the 
service connections for the maximum 
load to be demanded at any one time, 
and such apparatus in the generating 
and substations as would be equiva- 
lent to the private plants disabled at 
any one time. It is not necessary to 
duplicate the total installations of the 
private plants, 'but merely such a por- 
tion as would be at any one time out 
of use. 

To supply the second class of con- 
sumers, says the report, the supply 
company would need to provide the 
necessary service connections and dis- 
tribution lines for each user, "but as 
current would be demanded when the 
stations of the company would other- 
wise not be operated at their maxi- 
mum capacity, it would not be neces- . 
?ary to provide station equipment 
specifically for this service, but it 
might be necessary to keep certain 
units revolving slowly, but in this 
respect the service would not differ 
from the ordinary service, where there 
must be constant readiness to serve 
somewhat in excess of the demand at 
that moment." 

"The third kind of service," the 
report continues, "is the most expen- 
sive of all, for if any restriction were 
placed on the amount of current con- 
sumed or the time of its use, break- 
down consumers could easily so plan 
their equipment that they would have 
sufficient capacity to supply all the 
current required, except during the 
peak hours of the day. When the 
peak load began to come on, the break- 
down user could take sufficient cur- 
rent from the supply company to 
handle this peak, breaking the connec- 
tion when the peak hour passed and 
the consumption again became normal. 

"Inasmuch as the peak period for 
the breakdown consumer is apt to be 
the peak period for the supply com- 
pany this would mean that the com- 
pany would be called upon for a con- 
siderable amount of current during the 
short period when the demands of the 
ordinary consumers are at their height. 
Thus the supply company would be 
obliged to maintain a considerable 
plant during the whole day in order to 
supply the breakdown consumer dur- 
ing one or two hours and to carry the 
surface plant throughout the year in 
order to handle the winter maximum 
due to the short period of daylight and 
cloudy weather." 

Commissioner Maltbie finds that the 
public supply companies, inasmuch as 
they have franchises for the use of the 
streets, are under obligations to pro- 

vide breakdown service under ordi- 
nary circumstances, assuming that the 
rate is fair and that a reasonable profit 
is allowed. Prior to the passage of 
the law of 1905, fixing the maximum 
price of current in Manhattan at ten 
cents per kilowatt hour, the New York 
Edison Company supplied breakdown 
service to every applicant. Since that 
time, and until the Public Service 
Commission was created, the company * 
refused to supply breakdown service 
except to those who already had con- 
tracts. It was this refusal which 
brought the matter before the Public 
Service Commission and resulted in 
the investigation by Commissioner 

The Edison Company first offered 
to supply this breakdown service for 
the maximum charge or guarantee of 
$30.00 per kilowatt of installation. 
"As a result," says Commissioner 
Maltbie, "the individual who had 
lamps, motors and apparatus of a total 
capacity of 100 kw. would be obliged 
to guarantee an annual payment of 
$3000.00, no matter what amount of 
current he consumed ; no matter 
whether he used only 20 or 30 kw. 
at any one time." 

Commissioner Maltbie concludes 
from this that a charge based upon 
installation and not upon the maxi- 
mum demand would be generally un- 
just. After several conferences the 
representatives of the Edison Com- 
pany agreed to change this charge so 
as to base it on the maximum demand 
instead of on the total installation of 
private consumers. The breakdown 
rate as now agreed to, therefore, will 
be a service charge of $30.00, includ- 
ing the supply of electric light at the 
best rate of the class for each kilowatt 
of maximum demand, for which the 
consumer may make written request to 
the company, and which maximum 
demand is to be the basis of a year's 
contract. The company puts this rate 
into effect for one year, with the privi- 
lege of withdrawing it at the end of 
that time if it is found to be unsatis- 

American Circular Loom Co. 

The Chelsea plant of the American 
Circular Loom Co. was entirely de- 
stroyed by the recent conflagration 
which devastated Chelsea. The Com- 
pany opened offices the next morning 
in the International Trust Building, 
Boston, and began the work of as- 
sembling machinery for the manu- 
facture of circular loom, which was 
the only product manufactured at the 
Chelsea plant. 

The electroduct factory operated by 
the American Circular Loom Co., at 
Kenilworth, N. J., is_ in full operation, 
and the new metal molding plant, lo- 



June, 1908 

cated also at Kenilworth, is now fully 
completed, including a large electro- 
galvanizing plant. 

A new factory for the manufacture 
of circular loom has been secured in 
North Cambridge, Mass., and in a 
very short while circular loom will 
be shipped from the new plant. The 
most serious aspect of the fire was the 
very severe loss suffered by the em- 
ployees of the company, nearly all of 
whom lost house, home and all they 
possessed. One fatality resulted. 

The Copper Hand-BooK 

Volume VII of the new edition of 
the Copper Handbook has just been 
issued by the author, Horace J. 
Stevens, of Houghton, Mich. The 
book has 1228 pages, octavo, being 
materially larger than before. The 
author apologizes for his inability to 
revise the book throughout, explain- 
ing that fire, sickness and loss of five 
months' time prevented, but the new 
volume contains about 180,000 words 
of new matter. 

• The new edition of the Copper 
Handbook contains 25 chapters, an 
increase of nine, treating of copper 
under the headings of history, 
geology, chemistry, mineralogy, min- 
ing, milling, concentrating, hydromet- 
allurgy, pyrometallurgy, electrometal- 
lurgy, alloys, brands, grades, uses, 
substitutes, terminology, geography, 
copper deposits, copper mines and 
statistics. Price $5.00. 

Large Railway Motor Contracts 
to General Electric Co. 

Contracts made by the Chicago 
Railways Co. this year and by the 
Chicago City Railway Co. two years 
ago with the General Electric Co. for 
quadruple 40 h.p. equipment involved 
a cut of more than 20 per cent, by the 
industrial concern under regular 
prices quoted for such equipment. 
The former has ordered between 300 
and 400 cars and the latter has 
equipped about 700. There will be 
800 additional cars on the Chicago 
Railways lines within three years, 
bringing the total new equipment of 
both companies up to about 2000 cars. 

The two companies and the city 
saved about $1,000,000 by the bar- 
gains at which this equipment has 
been secured and contracted for. 

The General Electric Tungsten 

The General Electric Company has 
obtained very satisfactory results with 
the tungsten lamp. They have shipped 
over 75,000 to all parts of the country 
and the breakage in shipment is be- 
low one and one-half per cent. They 
state that they are issuing a new bulle- 

tin covering tungsten lamps, both 
series and multiple, and have a large 
production, good stocks and are in 
position to make prompt shipments, 
particularly of the 100-watt and all 
types of tungsten lamps which they 
have standardized. 

After complete tests made at the 
photometric laboratories of Purdue 
University, the Westinghouse four- 
ampere metallic-flame arc lamp has 
been selected bv *he Board of Works. 

"Westing'House Turbines for Manila 
and Japan 

No less than ten machines, aggre- 
gating 25,000 h.p., are included in a 
large shipment of Westinghouse 
turbo-electric power equipment from 
East Pittsburg to the Far East. Most 
of these machines will go to Japan for 
the equipment of railway, lighting and 
manufacturing plants. 

One of the first machines to be put 
in service will be a 1500 kw. turbine 
unit for Manila, to be installed in a 
station with four other machines of 
like construction put in service several 
years ago. This railway system was 
engineered and constructed by the 
American engineering firm of T. G. 
White & Co. 

Hardly second in importance is the 
large turbine station of the Osaka 
Electric Company, Osaka, Japan, now 
building. This will be one of the 
largest power stations in Japanese ter- 
ritory and will contain for the pres- 
ent 15,000 kw. in five units. Three 
of these machines are now being 
shipped from East Pittsburg. The re- 
mainder will follow as fast as they can 
be built and tested. The Osaka instal- 
lation is under direct charge of 
Messrs. Takata & Co., of New York 
and Tokyo. 

In the strictly manufacturing field. 
there are two installations in process 
of erection, for the Imperial Steel 
Works of the Japanese Government 
and the shipyards of the Hakkaido 
Tanko Steamship Co. Two 500 kw. 
Westinghouse-Parsons turbo units 
will comprise an initial installation in 
each of these plants. 

Low Lig'Hting' Rates For 
Lafayette, Ind. 

The city of La Fayette. Ind.. has just 
closed a contract with the Merchant-" 
Electric Light Association for light- 
ing the streets, alleys and public 
buildings for a period of 10 years. 

For the last 20 years the city has 
been paying $66 per lamp year, for 
2200 hr. of lighting, with 6.6-ampere 
direct-current open arc lamps. 

Under the new contract, which goes 
into effect Sept. 1, 1908, the company 
will furnish and light 300 metallic- 
flame arc lamps 2500 hr. for $37.98 
per lamp per year, and extra lighting 
at one cent per lamp per hour. This 
will amount to a saving of about $100.- 
000 to the city in 10 years. 

The engineerrng work, preparation 
of specifications J^ind contracts has 
been done by J. \V%lter Esterline, con- 
sulting engineer. 

"Williamsburg Bridge Cables 

Dossert & Company, 242 West 41st 
Street. New York, have designed a 
special extension fitting for their sold- 
erless cable taps, which are used as 
equalizers on the feeder cables supply- 
ing power to Brooklyn Rapid Transit 
trains which are to run over the new 
loop connecting the Williamsburg and 
Brooklyn Bridges. The cables range 
in size from 2.500.000 cm. to 500,000 
cm. The work is under direction of 
Latev & Slater, consulting engineers. 
The Gore Engineering & Contracting 
Company have placed orders for 244 
sets of these large double taps, 164 
regular taps and a number of two-way 
Dossert joint-;, including 134 connect- 
ing aluminum to copper. 


Mr. Wynn Meredith has become a 
partner in the firm of Sanderson & 
Porter and will have charge of the 
Western office which they have opened 
in the Union Trust Building, San 

After a technical training at the 
University of Illinois, Mr. Meredith, 
in 1888, became engaged in the con- 
struction and operation of lighting 
and railway properties. He was ac- 
tively connected with the engineering 
and operation of the electrical plant 
of the World's Fair at Chicago, in 
1893, and the California Fair, in 1894, 
subsequently becoming associated with 
Messrs. Hasson & Hunt, and later a 
member of the firm of Hunt, Dillman, 
Meredith & Allen. San Francisco, Cal. 

During fifteen years' residence in 
California Mr. Meredith has been en- 
gaged in general engineering work 
and prominently identified with many 
of the important hydro-electric and 
transmission developments on the Pa- 
cific Coast, in the United States and 

Announcement is made of the ap- 
pointment of Professor C. F. Harding 
as head of the school of electrical 
engineering of Purdue University. 
Professor Harding is a graduate of 
Worcester Polytechnic and has had a 
broad practical training as an engi- 
neering teacher. His special training 
has been along the line of high tension 
railway work, and he was electrical 
engineer for the first railway of that 
character in New England. He has 
(Continued on page 12.) 


Volume XXXIX. Number 6. 
$ 1 .00 a year ; 1 5 cents a copy 

New York, July, 1 908 

The Electrical Age Co. 
New York and London 

Published monthly by The Electrical Age Co., 45 E. 42d Street, New York. 

J. H. SMITH. Pres. C. A. HOPE, Sec. and Treas. 


Telephone No. 64J8 38th. Private branch exchange connecting all department!. 
Cable Address — Revolvable, New York. 


United States and Mexico, $1.00; 

Canada, $1.50. To Other Countries. $2.50 


Insertion of new advertisements or changes of copy cannot be guaranteed for the following 
issue if received later than the 15th of each month. 


Business and Editorial Departments, 33 Bedford Street 

Strand. W. C. 

London Telephone, Post Office, 3807 Central. Private branch exchange. London Cable 

Address — Revolvable, London. 
All remittances to London Office should be crossed London and County Banking Company. 


Subscriptions may be sent to the following branch offices or agencies, where they will re- 
ceive the same careful attention as at the home office in New York : 

124, Queen St., Melbourne 33, Loop St., Cape Town 

359, George St., Sydney Unter den Linden. 5, Berlin 

61, Main St., Yokohama Nevsky Prospect, St. Petersburg 

23, Esplanada Rd., Bombay 31, bis rue de Fauburg Montmartre, Paris 

General Agents for United States and Canada : The American News Company 

Returning Prosperity 

The electrical industry is still 
largely of a sort to feel the sensitive 
trend of the money market in spite of 
the fact that its manufactured prod- 
uct is a staple necessity, the demand 
for which should be as steady as that 
for oil, soap, sugar or other articles 
needed in daily life. The reason lies 
in the fact that the total electrical 
product is of two distinct kinds : First, 
the manufacture of supplies and ap- 
paratus to replace what is used up or 
worn out — and this business increases 
in poor times ; and second, the manu- 
facture of apparatus for new electrical 
projects which require to be financed. 
In a season of hard money the latter 
class of business becomes very light. 
It shrunk in the late panic to almost 
nil, falling to a point well below that 
reached in the depression of 1903-4. 

In a time of depressed trade, the 
supply business actually increases, 
owing to the fact that motors and 
other apparatus which in ordinary 
times would be scrapped are made to 
do duty even at the cost of excessive 
repairing. Were it not for the steadi- 
ness of this kind of business it would 
be well-nigh impossible for the large 
manufacturers to make a profit on 
their business as a whole. These facts 
were abundantly demonstrated in 
1904, which was the first, severe de- 
pression since the electrical industry 
had grown to a large volume. 

Conditions are not different at the 
present time, except that the recovery 
promises to be more rapid than usual. 
The large operating companies are 
raising money under bond issues and 
coming into the market for big ap- 
paratus. The demand for estimates 
on new work is now practically as 
heavy as ever. Collections are more 
satisfactory than in any previous de- 
pression. There are abundant indica- 
tions that the present depression in 
the trade has turned and that business 
is rapidly working to a normal level. 
This is the general opinion of well-in- 

formed men in the trade. In a very 
recent interview President Walter H. 
Whiteside, of the Allis-Chalmers Co., 
gave forth a statement which, while 
it applies to his company in particular, 
is in a measure true of the entire, in- 
dustry. He said: 

"There is a marked improvement in 
general business. Within the last 90 
days the bookings of machinery orders 
have increased from 30 per cent, to 
50 per cent, of normal. The business 
for June to date indicates that at 
least 65 per cent, of normal will be 
reached. Xew inquiries received are 
constantly increasing in number and 
volume. They are of a much more 
substantial character than at any time 
since the depression began. Their 
general tone denoted extensive plans 
in contemplation and for early de- 
velopment ; the question of prompt 
shipment being already regarded an 
important consideration. 

"The immediate requirements for a 
larger volume of new machinery for 
necessary extensions and improve- 
ments to existing plants are greater 
than for four years. This for the 
reason that with many operating com- 
panies purchases which under ordi- 
nary circumstances would have been 
made a year or eighteen months ago, 
were put off, due to the monetary and 
other conditions, until to-day the factor 
of safety in these plants is becoming 
a serious matter and new equipment 
must be added. 

"Unlike previous depressions, fol- 
lowing which collections for compara- 
tively long periods continued very un- 
satisfactory, there was a substantial 
improvement within a short time after 
the recent flurry, and at the present 
time collections are exceptionally 

"While some difficulty may still be 
experienced in the way of financing 
entirely new undertakings, yet the 
financial condition of the relatively 
small buyer is on a sound basis and 
credits are about normal. In fact, the 

improvement in new orders placed is 
from the smaller class of purchasers, 
who are, commercially, an important 
factor. Of course, this class of busi- 
ness is widely diffused, but in the ag- 
gregate it is greatly stimulating im- 
proved trade conditions. 

"The business outlook is distinctly 

"WestingHouse Readjustment 

is Success 

At the meeting of the readjustment 
committee of the Westinghouse Elec- 
tric Mfg. Co., June 26th, it was decided 
to give the merchandise creditors' 
committee and the stockholders' com- 
mittee until September 1st to put 
the finishing touches on their work. 
The readjustment- committee accepts 
the plan of the other committees but 
will allow a postponement in terminat- 
ing the receivership until September 
1st in order to thoroughly finish the 
work which was hastily done in the 
last few days before the conferences 
which resulted in the acceptance of the 
plan of the merchandise creditors' 
committee. George Westinghouse is 
to be congratulated for the splendid 
fight he has made. Much credit must 
also be extended to Jos. W. Marsh, 
chairman of the creditors' committee, 
L. A. Osborne, second vice-president 
of the company, and C. P. Humphrey, 
of the employees' committee for suc- 
cessfully getting the readjustment plan 
to a solid basis. 

While no public statement as to the 
amount of subscriptions has been 
made, it can be stated that it falls 
very little short of the stipulated 
$10,000,000 which the merchandise 
creditors set out to raise on April 2d, 
after the Jarvie readjustment com- 
mittee had wrestled with the problem 
for five months without evolving a 
plan which promised success. 

The fruition of the present work 
will enable the company to start busi- 
ness with quick assets approximating 




July, 1908 

$32,000,000, of which $12,000,000 
represents actual cash. The company 
will also have about 1500 new stock- 
holders, many of whom are influential 
in the trade. 

The Transformer Station 

The layout of switching devices in 
transformer stations has assumed an 
importance in late years of hardly less 
degree than that of the generating 
station itself. In main essentials the 
problems which develop are not dif- 
ferent from those which confront the 
engineer in designing the switching 
arrangements in the main power sta- 
tion. Failure to recognize this truth 
clearly, coupled with failure to handle 
the current in the same complete man- 
ner, has led to many costly lessons. 
Experience has shown that it is wisest, 
if not less costly, to take all pre- 
cautions against the development of 
arcs, their spreading to apparatus not 
directly involved, and the provision of 
means for automatically cutting out 
damaged conductors and apparatus. 
Every effort is now subordinated to 
the end of maintaining continuous 

In a paper read before the recent 
National Electric Light Convention 
at Chicago, and reprinted in full at 
another page, Mr. Stephen Q. Hayes 
discusses these matters in some detail 
and dwells particularly upon the 
features which provide automatic pro- 
tection. Mr. Hayes points out the 
difficulty of preventing an overload or 
short circuit occurring on one line 
from tripping all of the breakers on 
the other incoming lines, and indicates 
the proper selection of relays to avoid 
this trouble. 

The Copper Situation 

When Americans were buying cop- 
per last year around 25 cents Europe 
was keeping out of the American 
market by drawing upon her reserve 
stocks. Now Europe is buying very 
largely in the American market. The 
supply, however, is going largely 
into current production. There is no- 
where a visible large stock of copper. 

For the past four months the price 
of copper has been hovering between 
I2V2 and 13 cents, and while inquiries 
have increased very much, the buying 
is still only in moderate quantities. 
The Westinghouse and General Elec- 
tric Companies are reported as buying 
only for • present needs. The Allis- 
Chalmers Company is known to have 
placed a very large order, sufficient, 
it is said, to provide for their entire 
output during the coming year. 

The feeling in the trade is that we 
have passed the low point of the de- 
pression. In support of this opinion 
there is, first, an entire absence of cop- 

per stock with producers and con- 
sumers ; second, a normal state of 
business confidence based upon the 
clearing political situation ; third, the 
actual demand for copper is increas- 
ing ; fourth, the large operating com- 
panies are getting bond issues, the 
proceeds of which will later be used 
in plant extension. 

Daniel Guggenheim, president of 
the American Smelting and Refining 
Co., is reported as saying that it is 
now possible to sell all of their copper 
at 13 cents a pound, whereas four 
months ago it was hard to make sales 
at I2 1 - 1 cents. 

The N. E. L. A. Gas Engine 


The report of the X. E. L. A. Com- 
mittee on gas engines is voluminous. 
It covers 170 pages and represents a 
good deal of painstaking work. Ex- 
actly 13 pages present the experience 
of users of gas engines in ^^ different 
plants. From the list of installations 
furnished by the manufacturers for 
their investigation the committee has 
eliminated from consideration "all 
units of less than 300 h.p." 

While the word "unit" apparently 
refers to a single engine unit, still a 
reference to the tables proves this not 
to be the case. What the committee 
means is that it has eliminated from 
consideration every gas-engine plant 
having a capacity of less than 300 h.p. 
Some of the gas engines listed in the 
tables are under 100 b.h.p. 

The remainder of this bulky report 
is filled with a description of manu- 
facturers' descriptions of their prod- 
ucts — a most useful piece of work 
since this information is brought to- 
gether under the compass of a single 

Of the ^^ gas-engine installation^. 
18 run on natural gas and 12 on pro- 
ducer gas, though it is not recorded 
what type of producer is used, which 
is a highly important bit of informa- 
tion. It were also a desirable thing if 
the committee had indicated which 
engines were operated in the central 
station field. The total horse power 
installed as reported by the manu- 
facturers is 183,113 h.p., with 55,225 
h.p. generated by producer gas, of 
which 35,625 is from bituminous coals 
and 12,250 from wood and lignite. 
This fact is surprising in view of the 
generally reported unsatisfactory ac- 
tion of the bituminous producer. Of 
the ^^ plants reporting, 19 contain 
Westinghouse engines, five have 
Crossley engines, four Snow engines 
and three are of Koerting manu- 
facture. Definite replies as to the 
maximum length of time the gas en- 
gine can be kept in operation run any- 
where from a week to six months. 

Very few report trouble with igniters ; 
practically none are bothered with 
back-firing or pre-ignition. The con- 
sensus of opinion regarding reliability 
is that it is as good as the steam 
engine. Fourteen of them report 
parallel operation as satisfactory, 
while only one reply is unfavorable. 
The committee may be congratulated 
in confirming what has been very gen- 
erally understood — that the gas en- 
gine is a satisfactory piece of mech- 
anism. It is to be regretted that it 
did not obtain "results" in its investi- 
gation of the 31 producer plants which 
it enumerates. 

The report presents an impressive 
list of Westinghouse installations of 
300 h.p. or over, including 16 central 
station plants. There are about 
118,000 h.p. of gas engines built and 
on order by the Allis-Chalmers Co., 
which is a splendid showing, consider- 
ing that this company entered the gas 
engine field long after the Snow- 
Steam Pump Works and the West- 
inghouse Company. The total of 
Snow Gas Engines built and under 
construction is 135,510 h.p. 

The real question which should 
have engaged the attention of this 
committee is the producer end of gas 
power generation. It is to be hoped 
that next year the committee will have 
enough satisfactory data on this sub- 
ject to remove the current impression 
that the gas producer is uncertain in 
its action, and as yet largely in the ex- 
perimental stage in this country in its 
operation on bituminous coal. It 
would be advisable to collect some 
data on the initial cost and mainte- 
nance of stations of various size. We 
are of the impression that in spite of 
the halving of the full bill by the gas 
engine, that the maintenance of a gas- 
engine-producer outfit is nearly twice 
as much as a steam plant, that the in- 
itial cost of a small plant is about 50 
per cent, above that of a steam plant, 
and that with any size of station the 
operating force must be larger. 

Motors for Steel Mills 

For a long time motors installed in 
steel mills were used only for driving 
pumps, compressors, cranes and con- 
veying apparatus. Xo one was able to 
guarantee successful apparatus if mo- 
tors were applied to the main operation 
of rolling steel. The rolling mill offered 
about as severe conditions to overcome 
as one could find, the load varying 
from a few hundred horse power to 
20,000-30.000 at a maximum on the 
blooming-mill roll. The blooming 
roll takes a red-hot ingot about two 
feet by two feet by six feet, and passes 
it through the rolls back and forth, 
reducing it to the proper size for sec- 

July, 1908 



ondary rolling into building beams, 
rails or other finished shapes. 

So severe are the loads, jumping 
from a few horse power to hundreds 
or thousands in an instant, and going 
off equally sudden, that direct roll- 
ing by motors has been one of the last 
applications in displacing the steam 
engine. Engines for this work had been 
uneconomical, heavy, costly and sub- 
ject to continual repair. They were 
clumsy for the operator to handle, and 
in nearly every way undesirable. 

The chief reason, however, for 
wishing to displace the steam engine 
was that after a breakdown of the en- 
gine it might require anywhere from 
a half-hour to twenty-four or more to 
effect the repair, whereas, in the case 
of a motor, it is the work of only a 
few minutes to slip in the spare arma- 
ture or field coil, or even to replace the 
entire motor. The importance of this 
is realized when one reflects that one 
hour's shut-down from continual roll- 
ing at the Edgar Thompson rail mills 
means a loss of something like over 

Good rewards were offered manu- 
facturers if they could step into the 
rolling-mill field, and, finally, they 
have done so. Motors up to 10,000 
h.p. have been installed, and very 
probably the 25,000-h.p. engine now 
being built by the Allis-Chalmers 
Company will be duplicated by them 
with a motor of equivalent or greater 
rating when the time comes. 

In the solution of this problem, the 
natural first step was to add a heavy 
fly-wheel to ease off sudden changes 
in load. A second step was to use a 
very liberal rating for the normal ra- 
ting of the motor, while a third was to 
provide for excessive overload capac- 

ity amounting usually to 300 per cent 
for a half-hour. Rapid reversing is 
another point which had to be looked 
out for, and, therefore, armatures of 
small diameter had to be designed to 
reduce the so-called centrifugal effect. 
This meant a minimum total fly-wheel 

Both alternating and direct-current 
motors have been used when direct 
current they have been compound 
wound on the basis of 15 per cent, 
compound. 85 per cent, shunt. 
When of the alternating-current 
type, the rotors have been of the 
wound construction, so that extern- 
al resistance might be inserted for 
large starting torque, or speed con- 
trol, or both. 

The commonly used voltages are 
230 when direct-current equipment is 
used, and 220 or 440 for alternating 
current. A frequency of 25 cycles 
only has been used on account of the 
slow speeds to be dealt with on the 
rolling-mill shafts. 

Smaller motors are built of the en- 
closed type, while from about 150 h.p. 
open motors are used. In the latter 
case the motors and controlling appa- 
ratus (except for the drum-type con- 
troller used by the operator) are in- 
stalled in a room of brick or sheet 
metal with the shaft extending 
through the wall. Frequently, how- 
ever, the cover is only a large box 
which may readily be removed from 
the motor. Mill work is a special 
field, and each problem has to be con- 
sidered separately on its own merits. 

Testing Lamps by Substitution 

In making photometric tests of 
incandescent lamps two methods are 
available: The lamp to be tested may 

be compared either directly with a 
standard lamp, or with one of a series 
of lamps which have been thus 
standardized. The direct comparison 
method has the disadvantage of burn- 
ing the standard lamp during the en- 
tire test and thus shortening its period 
of usefulness. For this reason it is 
not recommended for most purposes, 
and the substitution method is pre- 
ferred. This method has the advan- 
tage of automatically illuminating 
photometric and instrumental errors, 
besides those due to the personal 

In testing lamps by this method 
care should be taken that good con- 
tact is secured between the lamp base 
and its holder. It is also important 
to connect the voltmeter leads as near 
as possible to the lamp terminals. Be- 
fore closing the standard lamp cir- 
cuit, care should be taken to avoid 
the possibility of burning the lamp at 
a higher voltage than that at which it 
is rated. When the circuit is closed 
the voltage upon the standard lamp 
should be gradually increased until 
its standardization voltage is reached. 
The current flowing through the lamp 
should then be determined. If under 
these conditions the current through 
the lamp is more or less than that 
specified in the standardization cer- 
tificate, then either the meters are in 
error or the standard lamp has 
changed. Reference to the other 
lamps of the standard series will then 
indicate which is the case. If the 
standard lamps are in good agree- 
ment among themselves, and the dis- 
crepancy still exists, one of the meters 
may be assumed correct and a cor- 
rection applied to the other sufficient 
to bring the wattage up to the stand- 

Some Points to be Considered in the Purchase 

of Steam Turbines 

PROSPECTIVE buyers of steam 
turbines are usually at a loss to 
know just what features of the 
steam turbine merit consideration. 
The representative of an impulse type 
is liable to point out the defects of the 
reaction turbine, but remain eloquent- 
ly silent on those features of his own 
make which he desires to remain un- 
discussed. The relationship is re- 
versed if a reaction man has the field 
to himself. Some of the features of 
turbines which generally come up in 
turbine negotiations are given here in 
order that our readers may be better 
prepared to receive the turbine sales- 


Impulse turbines are in almost every 
respect similar to Pelton water-wheels 
in principle. Pelton water-wheels are 
used only under high heads and great 
velocities. The buckets of impulse 
steam turbines operate likewise at 
high velocity by reason of the steam 
speed. As the steam passes from the 
boiler to the turbine, its speed is rela- 
tively low. Just before reaching the 
buckets it ordinarily passes through 
expanding nozzles, which allow its 
pressure to drop and volume to in- 
crease. Evidently it will have to dis- 
charge at a higher rate than it entered 
the expanding nozzle in order to make 
way for the high pressure steam fol- 
lowing. It is in this way that the 
potential energy of the steam (due 
to its pressure) is changed into kinetic 
energy (energy of motion), causing 
the steam to impinge on the buckets 
at enormous velocities. 

The reaction turbine is similar to 
the well-known Samson or the Esher- 
Weiss water-wheel. The former are 
well known when low heads are to be 
used. The latter type of wheel is used 
for medium heads and is installed at 
the original Niagara power plant. 
In the reaction type of steam 
turbine, the steam enters the turbine 
at boiler pressure and expands 
through stages down to atmospheric 
pressure, or into a condenser. 

Both types have equally enthusi- 
astic supporters, and for straight 
steam economy there is practically no 
choice, where conditions remain nor- 
mal, manufacturers' claims to the con- 
trary notwithstanding. 

The best-known impulse type is the 

Curtis machine, manufactured by the 

General Electric Co. Up to 300 kw. 

it is of the horizontal type, but of 



vertical forms in sizes from that point 
up. If .necessary, it can be built in 
larger sizes for the horizontal position. 
The YVestinghouse Company first 
made the reaction turbine popular in 
the United States. They have always 
made them of the horizontal type only. 
This same type of turbine is manu- 
factured by the Allis-Chalmers Com- 
pany, which has added a number of 
features of design of its own develop- 


In considering the relative merits 
of the two types of turbines, it is 
found that the impulse type has the 
slower shaft speed, in some cases by 
one-half. This, too, in spite of using 
higher steam velocities. The shaft 
speed is slower because the buckets 
are mounted upon the rim of large 
diameter disks, giving high peripheral 
velocities (up to five miles per min- 
ute), but quite slow for turbine shaft 
speed. In the reaction type the blades 
are mounted upon an enlarged section 
of the shaft itself, giving a more com- 
pact size but a higher shaft speed. 

The question naturally arises, why 
not use large diameters on the reac- 
tion type and obtain slow speeds? 
The answer is, because the steam leak- 
age (principally) would be increased. 
In this type the steam is expanding as 
it traverses the buckets, and conse- 
quently has motion in every direction. 
If the diameters were increased, it 
would mean increased leakage area be- 
tween the stationary and the revolving 
part. In the impulse type, on the con- 
trary, the steam is practically all ex- 
panded in the nozzles before it gets 
to the interior, and therefore has 
motion in one direction only. Con- 
sequently, an increase of clearance 
has less effect. 

This fact became evident only after 
much experimenting. 

In both types of turbine, clearance 
between moving and stationary 
buckets is of the same relative im- 
portance, while in the impulse type 
the radial clearance is not a factor 
in the economy of the machine. 
The radial clearance is of import- 
ance in the reaction type. 


It was warping in the outer shell 
which led to early troubles in the 
Parsons type. The first big im- 
provement of the Allis-Chalmers 
Company was to place a channel 

iron on the ends of the blades. 
While warping might cause the 
outer shell to wear into this iron ; at 
least, the operation of the turbine 
was not interfered with. The West- 
inghonse Company later adopted a 
method of strengthening its biades 
by drilling a hole close to the end 
and threading a comma-shaped wire 
through it. This peculiar cross-sec- 
tion was selected so that the edge 
might be turned down between the 
blades to insure proper spacing and 
non-interference with the flow of 


The impulse type needs no balanc- 
ing for end thrust, as the steam enters 
each bucket and leaves it at approxi- 
mately the same angle. 


Consequently the pressure is always 
in the direction of rotation. In the 
Parsons type two forces appear : one 
in the direction of rotation and the 
other parallel to the shaft, and added 
to this we have unbalanced steam 
pressure in the rotor. . To offset this, 
pistons are mounted at the emission 

Fteej^y ge 


end of the shaft of sufficient area to 
counterbalance the end thrust. Live 
high pressure steam is admitted to 
the chamber containing them to fur- 
nish the counterbalancing force. The 
Westinghouse Company is also build- 
ing large sizes of the double-flow type, 
which is self-balancing. 


In the Curtis turbine all buckets 
are of a uniform size and of about 
three times the cross-section of the 
blading of the Parsons turbine. The 
Parsons' blading increases in size as 
the steam passes toward the low-pres- 
sure end. The theoretically perfect 
shape of a Parsons rotor would be a 

July, 1908 



cane. Commercial and mechanical 
reasons, however, limit the changes in 
blading to stages. There has been 
some discussion as to the methods of 
fastening in the buckets and blades. 
But after much experimenting on the 
part of the manufacturers, methods 
have changed but little. With the 
continued use of the turbine it can be 
safely said that this point has but lit- 
tle real weight as a cause of trouble, 
or a hindrance in case of repair. 
Curtis buckets have a ring on the out- 
side which serves as a guide to keep 
the steam from spilling from the buck- 
ets, and to hold them together. In 
the use of the words "bucket" and 
"blading," both terms are employed 
to perform the same function in the 
turbine, but impulse type machines 
have the receptacle or cup called a 
"bucket," while in the reaction ma- 
chine, on account of the smaller size 
in cross-section, the word "blade" is 


In the Parsons type an improved 
throttling governor is used. This re- 
duces the volume of steam admitted to 
the turbine with change of load. It is 
evident that while steam is admitted 
to the high-pressure end of the turbine 
for ordinary work, there is no reason 
to prevent high-pressure steam from 
being turned directly into the second 
stage. This is a feature which the 
makers of the Parsons turbine offer 
with their machine, and, by it, enor- 
mous overloads can be carried, as the 
second stage blading has a much larger 
surface than the first stage. Of course 
the efficiency of the turbine is dis- 

turbed while it is operating this way, 
but as an emergency feature it is very 

The Curtis turbine is like a gas en- 
gine in that it has a definite load limit. 
This may be 25 per cent, or 50 per 
cent, more than the normal rating, ac- 
cording to the requirements of the 
buyer and his willingness to pay for 
it. Ordinarily the overload capacity 
is based on the overload capacity of 
the standard generator, which is 25 
per cent, for two hours. A given 
amount of blading with a definite vol- 
ume of steam at a predetermined pres- 
sure, vacuum and superheat can give 
but one maximum horse power. If 
this is called five-fourths of normal 
load, then 25 per cent, overload is 
obtainable. If a greater bucket sur- 
face is used and this is six- 
fourths of a normal rating, then the 
turbine has 50 per cent, overload ca- 

With a given maximum horse 
power to deal with, the manufacturers 
divide up the port area of steam en- 
trance into six, eight or ten valve di- 
visions. If the load is light, one valve 
opens and the others remain closed. 
As the load increases the tendency to 
slow down, the speed causes a second, 
third or fourth valve to open. These 
valves are hydraulically operated and 
are controlled by an automatic device 
which will close them all if the turbine 
speed reaches a predetermined limit. 


In the Parsons type the combined 
turbine and generator unit is sup- 
ported by three babbited bearings of 

the pedestal type. In the Curtis ma- 
chine there is one bearing — the step- 
bearing. The vertical shaft carries a 
cast-iron shoe on its base, which re- 
volves with the shaft in close parallel 
position with a stationary plate of cast 
iron of similar form to the rotating 
plate. The weight of the rotating 
turbine and generator is carried on a 
film of oil between the stationary and 
rotating plates. There is sufficient 
pressure on the oil to remove all met- 
allic friction. With the vertical tur- 
bine, the oil pressure under the step- 
bearing is about 400 lb. per. sq. in., 
while with the horizontal turbine the 
oil pressure is only sufficient to insure 
positive lubrication. 


In addition to the two makes cited 
above there are several others now on 
the market which are well known. 

The Rateau, Terry, DeLaval, are 
all of the impulse type and differ from 
each other only in details of construc- 
tion. For example, the Rateau uses 
a special entrance system of nozzles 
which change in size with each stage. 
The Terry is like a Pelton wheel. 
The steam, after striking a bucket, is 
deflected back into the case which 
contains the live ^team nozzle, and is 
there turned back against another 
bucket. The DeLaval is a single- 
stage turbine using expanding noz- 
zles and consequently high speed on 
periphery and shaft. A 10-h.p. ma- 
chine with a 10-in. wheel has been 
built with a shaft speed of 25,000 rev. 
per min. The peripheral velocity is 
about 10 miles per minute. 

Horsepower Required for Metal 
WorKing Tools 

By M. G. Buckley. 

Many tests have been made on met- 
al-working tools, such as lathes, drills 
and the like, to determine the proper 
horse-power motor to apply. The re- 
sults have been varied, indeed, de- 
pending entirely on local conditions. 
A motor manufacturer has summar- 
ized these tests to determine the 
average motor required. The fol- 
lowing may be safely used for 
any ordinary installation. Where 
high-speed steel is used or some 
other condition enters into the 
question, then special tests must be 


22 and 24 in 2 h.p. 

26 to 30 in 2.y 2 

36 to 42 in 3^ 

48 to 54 in 5 

72 in 6 

84 in 8 

For forge lathes use motors 50 per 
cent, larger than above. 


26 x 26 in.-8 ft 5 h.p. 

30 x 30 in 10 

36 x 36 in 10. 

38 x 38 in 12 

42 x 42 in 12 

44 X44 in 15 

48 x 8 in 15 

54 x 54 in 15 

60 x 60 in 18 

72 x 72 in 22 

84 x 84 in 25 

96 x 96 in 30 

120 x 120 in 40 

144 x 144 in 60 


10-in. swing 12 h.p. 

12-in. " 14 

14-in. 14 

16-in. " 15 

37-in. 4 

51-in. " 5 

60-in. " 7 

72-in. " 7 y 2 

84-in. " 10 

96-in. " 10 


16-in. stroke 3 h.p. 

18-in. " s l A 

24-in. 6 

30-in. " ey 2 


10-in. stroke 4 h.p. 

14-in. " $y 2 

18-in. " 6y 2 

24-m. " 7I0 

30-in- 7V2 


20-in \y 2 h.p. 

24-in 1 • - 

30-in 3 

36-in 3 

40-in 4 

50-in 5 

60-in 5 


26-in. between housings 8 h.p. 
36-in. " " 12 

42-in. " " 15 

48-in. " " 15 

The Central Station Distributing System 

Transmission and Conversion 


IN the development of a distributing 
system, the radius of transmis- 
sion from the point of supply 
tends to increase as the population 
grows. After a time the number of 
feeders to certain districts remote 
from the generating station becomes 
such that the transmission may be 
effected at a higher voltage to much 
better advantage. Such transmission 
involves transforming and regulating 
apparatus at a point remote from the 
generating station, which in turn re- 
quires a building and other accessories, 
and the result is a substation. This 
substation involves an investment in 
real-estate (or a rental charge), trans- 
forming apparatus, switchboard, etc., 
and an operating expense for at- 
tendance and repairs. On the other 
hand, the feeders running into a dis- 
trict occupy valuable duct or pole 
space and require a large investment 
in copper and insulating materials. 
It, therefore, becomes profitable to es- 
tablish a substation when the amount 
required to pay fixed charges on the 
substation investment and its operating 
expenses is about equal to that re- 
quired to meet the fixed charges and 
maintenance expense on the feeder 
equipments which would be required 
if a substation were not installed. In 
a growing system it may be advisable 
to anticipate this point somewhat and 
install the substation earlier, in order 
to avoid the loss due to the installation 
and removal of feeders which are 
transferred to the substation after but 
a few years service. 

The point at which the balance be- 
tween substation cost and feeder cost 
is struck varies widely with different 
systems and . classes of construction. 
In a low tension direct-current under- 
ground system, the number of substa- 
tions is usually greater than in an al- 
ternating system with 2200 volt mains 
because of the shorter radius of action 
in low-tension systems. 

There are also many local condi- 
tions to be considered and two prob- 
lems are rarely, if ever, identical in 
every particular. With a given class 
of construction, the radius of distribu- 
tion and therefore the number of sub- 
stations is fixed by the voltage of dis- 
tribution, and second by the load. 

Commonwealth" Edison_Co., '.Chicago 


With a feeder loss at maximum load 
of 10 per cent, the length of a feeder 
is approximately one mile for each 
1000 volts of feeder pressure. On 
this basis the radius of distribution at 
220 volts is 0.22 mile or 1100 ft., and 
at 2200 volts it is 2.2 miles. There 
are usually some feeders which are 
longer than this on which the loss runs 
higher. When these become sufficient- 
ly numerous an additional substation 
becomes necessary. 

It is sometimes necessary to estab- 
lish a substation on account of a large 
block of load, such as an amusement 
park, large retail store, manufactur- 
ing plant, or other similar enterprise. 

The selection of a system of trans- 
mission for the wholesale distribution 
of energy from the generating station 
to substation is a matter of great im- 
portance. The three-phase system is 
used almost universally for this pur- 
pose, owing to its inherent economy 
of copper, reduced cost of generating 
apparatus and its adaptability to rotary 
converter and motor generator work. 

The voltage employed in the trans- 
mission system should be high enough 
to permit the economical supply of 
the most remote sections of the city, 
and, if possible, should be capable of 
reaching suburban substations. This 
usually does not require a voltage 
higher than 13,060, which permits the 
use of generators wound for the trans- 
mission voltage without step-up trans- 

In the larger cities where the loads 
to be transmitted are very great in 
proportion to the distance, it is de- 
sirable to use a voltage high enough 
to keep the transmission cables within 
reasonable limits of size. Voltages up 
to 20,000 have been found desirable 
in some of the larger cities in con- 
nection with transmission to suburban 

W nere the major portion of the en- 
ergy generated is not distributed from 
the generating station but is trans- 
mitted to substations for distribution, 
the generators should be wound for 
the transmission voltage in order to 
save the expense of step-up trans- 

The voltage selected should be one 
which is standard with manufactur- 
ing companies in order to secure lower 

first cost, and to facilitate delivery 
of coils and other apparatus which 
may be required for repairs. 

In American practice, two fre- 
quencies are standard for transmission 
purposes, namely : 25 and 60 cycles per 
second. Other frequencies, such as 
30, 40 and 66 cycles, are in use to a 
limited extent, but are not considered 
standard. Twenty-five-cycle current 
is found preferable when the major 
part of the energy transmitted is con- 
verted to direct current for distribu- 
tion. This is so because of the fact 
that rotary converters are much more 
stable in their operation at the lower 
frequency than they are at 60 cycles. 

Where transmission is effected by 
underground lines the charging cur- 
rent of the cables at 60 cycles becomes 
an important factor in a large system 
and may result in high potential 
surges in the transmission system in 
connection with switching operations, 
synchronizing and disturbances, due to 
the occurrence of short circuits or 

Twenty-five-cycle current, how- 
ever, cannot be used for arc lighting 
and is not in general use for incan- 
descent lighting, except out of doors, 
owing to the noticeable flickering of 
the light. It is, therefore, customary 
to convert the energy to 60 cycles 
for distributing purposes where 25- 
cycle energy is used in transmission. 
Sixty-cycle motors and transformers 
are less expensive than 25 cycle, which 
further favors the use of the higher 
frequency for distribution. 

In a system embodying a number 
of substations and perhaps more than 
one generating station, the design of 
the transmission lines supplying each 
substation with reference to continuity 
of service becomes a matter of great 
importance. The larger and more im- 
portant substations should have at 
least two sources of supply, one of 
which should be a separate trans- 
mission line direct from the generat- 
ing station to the substation. The 
second line may be a tap from a line 
which acts as a reserve for two sub- 
stations. In some cases the line may 
be tapped at an outside point and in 
other cases the geographical arrange- 
ment permits one line to be looped into 
the substation nearest the power- 

July, 1908 


145 l 

house, so that the line from this sub- 
station to the farther one becomes a 
tie line. Such a tie line, when pro- 
vided with a suitable arrangement of 
switches and bus connections, forms a 
very desirable reserve supply for both 
substations, as it can be fed from 
either end. 

In a low tension direct-current sys- 
tem with storage battery reserve, the 

i i 

TVu Lint 20,000 Toitl -^-| 
u Kvaoflton 




\ Lombard 



— , — 

— . — _- - . — 

Hockwell ( 
C.4 O.P.F.'.. 


Mn. LI. 



smaller substations, which are op- 
erated only during the heavy load 
period, are sometimes so located that 
the expense of having reserve trans- 
mission line capacity is so great as not 
to be justifiable. 

The development of a portion of the 
transmission system in the City of 
Chicago up to 1908 is illustrated in 
Fig. 1. It will be noted that in this 
system there are two generating sta- 
tions producing 25 cycle, 9000-volt 
current and several subsidiary steam 
plants, which are operated only dur- 
ing the heavy load period of the win- 
ter months. This system is somewhat 
different from that of other large 
cities in that several of the direct-cur- 
rent substations are used exclusively 
for the supply of 500-volt current to 
street railway companies and elevated 
roads. It will also be noted that prac- 
tically all of this system is under- 
ground, there being approximately 270 
miles of three conductor 9000-volt 
cable in service. 

Substations may be divided into 
three general classes : 

A. Those delivering alternating 
current from static transformers. 

B. Those delivering alternating 
current from frequency changing mo- 
tor generators. 

C. Those delivering direct current. 


Q Steam Plant 

(3) Steam Plant and Sub Station 

@ Railway Sub Stations 

% Sub Stations Distributing 
Direct Current 

■ Sub Stations Distributing 
Alternating Current 

Fig 1. 

In the first class, those delivering through oil switches. The step-down 

alternating current from static trans- transformers supply to the distributing 

formers, the equipment is compara- bus the proper pressure and from this 

tively simple. The incoming trans- bus the outgoing feeders radiate after 

mission lines supply a high-tension bus passing through their potential regu- 



July, 1908 

lators. Ammeters, volt meters, light- 
ning arresters and other accessories 
which are necessary for the proper op- 
eration of the equipment and for rec- 
ord of the output complete the equip- 
ment. There being no moving ap- 
paratus, a single attendant at the 
switchboard can properly care for 
pressure regulation, make minor re- 
pairs and keep the equipment in con- 
dition in most cases. The transform- 
er units should be so selected that the 
failure of a unit will not remove too 
much of the capacity of the substation. 
A spare unit should be at hand which 
can be quickly put into service in 

this may become a serious difficulty, 
owing to the space required for the 
air ducts. The circulation of water 
or oil permits more rapid cooling, and 
is, therefore, desirable in the larger 
units in order to keep the first cost of 
the transformer within reasonable 

The outgoing feeders should be 
equipped with oil switches capable of 
opening under load and provided with 
overload relays set to operate them 
in case of short circuit. Transfer and 
tie switches and others which are not 
required to be operated under load 
may be of less expensive design. 

All instruments measuring high po- 

Qu~<Sinnrcn •-, 


Bus S Z 

Tq f/V3TJH/AI£MT& /tMV 

-* Jo frw7-/iLi»r£Arr3 Amo 
t TO £zc/7tZ* Dvj 

7b BxxrrtH Bus ♦ 
F>£J.Ji Mmd UncMaacS' 



to /tarm/rtE 

-AND O/EAiJ0/!0 : B£b.lirj 

Jo 'Ma-maMs/Tu atvo 

TOMZCH3 •— I *&_k 

I } «JI". il* 

Oil. £wrrcH£S 

4ZX>r A*u 

Fig 2. 

The transformer equipment may be 
of the air-blast type, oil-cooled, water- 
cooled, or of the more recently de- 
veloped oil-cooled type, in which the 
oil is circulated through the trans- 
former by means of a system of piping 
and an external pump and radiator 
which carries away the heat. 

Where overhead lines are used the 
oil-cooled transformers are less sub- 
ject to puncture by lighting or high 
potential surges. 

Air-blast transformers involve 
blowing apparatus and ducts for the 
fresh air supply. In a large substation 

tential energy should be supplied 
through transformers, the low po- 
tential circuits and transformer cases 
being carefully grounded to prevent 
injury to operators or construction 

In the second type of substation, 
the supply of alternating current for 
distributing purposes is derived from 
frequency changing sets. These con- 
sist generally of synchronous motors 
driving 6o cycle generators, motor and 
generator being mounted on the same 
shaft. Synchronous motors are found 
desirable for this class of work be- 

cause of the fact that the windings 
may be designed for the transmission 
voltage and the expense of transform- 
ers thus avoided. The control of the 
power factor is also a great ad- 
vantage. The 6o-cycle generators de- 
liver a pressure suitable for the dis- 
tributing feeders, and the switchboard 
and its equipment are therefore simi- 
lar to that employed in a transformer 

substation. The essential elements of 
such a substation are illustrated dia- 
grammatically in Fig. 2. The chief 
point of interest about such a substa- 
tion is the method of starting and syn- 
chronizing the motor-generator sets. 
When a unit is to be put in service 
it is transferred to a starting bus, 
which is supplied by an auto-trans- 
former which delivers about 40 per 
cent, of the transmission line pressure 
to the starting bus. The switches con- 
trolling the direct current for the fields 

Fig 4 

of the motor are left open. The oil 
switch controlling the motor is then 
closed on the starting bus and the unit 
begins to revolve as a hysteresis and 
induction motor. When the unit is 
at approximately synchronous speed 
the field is excited' and the unit is 
drawn into step as a synchronous mo- 
tor. This usually causes a rush of 
current for a very brief interval of 
time, as the machine may be out of 
phase at the time the fields are ex- 

When the conversion is from 25 
to 60 cycles this usually does not 
complete the operation of synchroniz- 
ing, as the 60-cycle generator is not 
necessarily in phase with its bus when 
the 25-cvcle motor has been synchron- 
ized. The ratio of field poles on the 
25-cycle motor to those on the 60-cycle 
machine must be as 25 is to 60 or as 
10 is to 24. When a 10-pole field is. 
mounted on the same shaft with a 24- 

July, 1908 



pole field, as is usually the case in a 
25-60 cycle frequency changer, only 
one set of poles on each field can be 
lined up in the same radial place. In 
Fig. 3 the poles which are aligned 
in the same radial plane are repre- 
sented by the heavy diameters. When 
the 25-cycle machines are synchron- 
ized, any of the five sets of poles on 
the incoming machine may fall into 
step with the poles represented by the 
heavy line on the operating unit. Fig. 
4 represents a unit in which the 25- 
cycle machine has fallen into step with 

Fig 5- 

a displacement of one pair of poles. 
This holds the incoming 60-cycle ma- 
chine out of phase with the operating 
unit, as shown by the dotted vertical 
line. If the rotation is counter clock- 
wise, the machines can be brought into 
phase by removing the supply of ener- 
gy from the incoming machine and 
allowing it to slip back one pair of 
poles until the heavy lines are in phase 
with each other. The machines are 
then in phase on both 25- and 60-cycle 
ends. The operation is carried out 

When the 25-cycle machine is 
locked in step by the excitation of its 
fields, the 60-cycle synchroscope 
pointer assumes one of five positions, 
as in Fig. 5. If it comes in on No. 1 
the oil switch is opened, cutting off 
the supply of power to the motor and 
it begins to slow down. When it has 
slipped back one pair of poles the 25- 
cycle synchronscope pointer will have 
made one revolution and the 60-cycle 
synchronscope 2.4 revolutions. This 
brings both synchronscope pointers 
into a vertical position and the oil 
switch is closed at this moment, lock- 
ing the motor in step again. The 
motor is now thrown to full pressure, 
and the 60-cycle generator connected 
to its bus. 

If in synchronizing the 60-cycle 
pointer had taken position No. 4, it 
would be necessary to "slip poles" 
four times before the 60-cycle ma- 
chine could be thrown in. 

These complications do not arise in 
synchronizing a single frequency 
changer with a 60-cycle generator 
driven by a prime mover, as the phase 
of the generator can be adjusted to 
any point by adjusting the governor 

The supply of direct current for the 
excitation of the fields of the motor 
and generator may be derived from 
direct-connected exciter sets or from 
separately driven exciter units, which 
may be used interchangeably with any 
motor generator. Where more than 
one unit is operated the use of 
separately driven exciters allows 
greater flexibility of operation, and 
is, therefore, usually preferable. 

Where direct current is used for the 
operation of auxiliary devices or for 
automobile charging, it is very im- 
portant to have separate exciter units, 
so that the variations of load due to 
auxiliary apparatus will not affect the 

units for this purpose. The switch- 
board containing the controlling ap- 
paratus for the exciter system and 60- 
cycle generators and feeders appears 
in the background. The vertical unit 
is carried on a step-bearing similar 
to that which has been developed for 
the Curtis turbine. Several of these 
units are in service in the City of 
Chicago. The special advantage in 
their use is a saving in floor space and 

The third class bf substation sup- 
plies direct current to a low-tension 
network through the medium of syn- 
chronous converters or motor genera- 

The electricity received from the 
transmission system passes through 
suitable oil-switching arrangements 
to step-down transformers, which de- 
liver a secondary pressure suitable for 
the rotary converter. From the trans- 
formers the current passes through a 

Fig 6. 

generator fields. A typical substation 
of this class is illustrated in Fig. 6. 
The two horizontal units are rated at 
1000 kw. each, while the vertical unit 
is rated at 2000 kw. The exciters 
for the horizontal units do not ap- 
pear, but it will be noted that the ex- 
citer for the vertical unit is mounted 
at the top, the armature being carried 
on the main shaft. In this substation 
the exciter for the 2000-kw. unit is 
provided with suitable switching ar- 
rangements, which permit its use as a 
direct-current motor in bringing the 
unit up to speed. Direct current is sup- 
plied from one of the other exciter 

potential regulator to the collector 
rings of the converter and thence 
through its windings to the commuta- 
tor from which direct current is de- 
livered to the brushes. The direct 
current passes through a circuit 
breaker and switch to its bus-bar, 
from which the feeders are carried to 
the distributing mains. A group of 
feeders may be isolated on one bus 
during the heavy load period and thus 
regulated separately. An ammeter is 
provided on each outgoing feeder and 
such other instruments as are required 
for a proper record of the output are 
installed on the converter panel. A 



July, 1908 





7i Ft£MO £>j^CHA-nos. 

^\T/£LJO J?£.S/3T*A/CS 

To + Bus, 

7i Ca.y-7/xu. 

*-*" I I Q- 

StCTlOn .. CORE •• «£ TNRf t-flttif TMIUrWl W 

Fig 8. 

cared for. The unbalance in a large 
system is rarely over 5 per cent, and 
the scheme is found very satisfactory 
in most instances. 

The use of the six-phase connection 
and converter winding reduces the 
length of the path traveled by the cur- 
rent in passing through the armature 
and thus reduces the losses and the 
heating. Theoretical calculations 
based on sine waves indicate that 
a direct-current generator rated at 
100 kw. may be rated at 131 kw., as 



* C.Mere, 


f v Y v 


1 V 


t- /— -t 


To Dj*£s 




Fig 7- 

voltmeter is required on the incoming 
transmission line. The essential ele- 
ments of a converter substation are il- 
lustrated in Fig. 7. 

The three-phase shell type of trans- 
former, air cooled, has been used quite 
generally for this class of service, 
owing to the economy in first cost and 
in floor space. The cross-section of 
the core is shown in Fig. 8. The 
middle section of such a wait must be 
connected up in a reversed sense in 
order to maintain the same magnetic 
density in each limb of the core. The 
air for cooling is blown through ducts 
within the case, and in substations of 
2000 kw. or more, it is sometimes 
necessary to provide ducts to carry the 
heated air outside the building. This 
may largely offset the saving other- 
wise effected. 

In a large three-wire Edison sys- 
tem it is desirable to use converters 
wound for the voltage across the outer 
wires, in order to avoid the multipli- 
cation of the number of units, and the 
increased expense incident thereto. 
The unbalance of the system may be 
cared for by one pair of no-volt ma- 
chines or by a motor-generator bal- 
ancer set, or by the use of six-phase 
diametrically connected transformer 
secondaries arranged as in Fig. 9. 
The latter plan has the great advantage 
that no 1 10- volt machines are required 
in the substation and that six-phase 
converters may be used with a greatly 
increased output from a given sized 
machine. The neutral of the direct- 
current system is connected directly 
to the secondary neutral of the trans- 
formers, and any unbalance is thus 

Fig 9. 

a three-phase converter, or at 194 kw., 
as a six-phase converter. 

The theoretical ratio of transforma- 
tion in voltage in passing from the col- 
lector rings on the alternating-current 
side to the direct-current brushes is, 
approximately, as 61 to 100 in a three- 
phase converter and as 71 to 100 in a 
six-phase converter. These are based 
on the assumption of a sine wave of 
electromotive force and may, there- 
fore, vary somewhat in actual prac- 

It will be noted that the converter 
is protected by a reverse-current relay. 
which opens the circuit breaker in case 
the* flow of energy is reversed, as in 
the case of a break-down in a trans- 
mission line, and shuts the machine 
down. Without such protection the 
reverse current might weaken the 

July, 1908 



fields of the converter and cause it to 
speed up quickly to a dangerous speed. 
The reverse-current relay does not 
usually operate below 20 per cent, of 
full load, and a speed limit, consisting 
of a centrifugal switch, is provided as 
further insurance against dangerous 
peripheral speeds. The speed limit is 
rarely called upon to act, and should, 
therefore, be tested at regular inter- 
vals. Accidents to converters in which 
machines have been wrecked have oc- 
curred in nearly all large systems, and 
the provision of such accessories must 
not be overlooked where the unit op- 
erates in parallel with a direct-current 
system having other sources of supply. 

As variation of the bus pressure by 
means of the field rheostat cannot be 
had without interference with the 
power factor in an ordinary synchron- 
ous converter, it is necessary to pro- 
vide a potential regulator. This is 
preferably of the induction type and 
is placed on the secondary side of the 
transformer between the transformer 
and the converter. This location being 
remote from the operator in many in- 
stances, the regulator is usually op- 
erated by a small motor controlled 
from the main board. 

Recently there has been developed a 
type of converter having split poles 
which are so designed that a con- 
siderable range of pressure regulation 
by means of the field rheostat is per- 
missible, without serious interference 
with the power factor. 

The arrangement of starting devices 
for synchronous converters is a mat- 
ter of great importance, as it must be 
possible to start them quickly and 
without serious disturbance to the sys- 
tem in regular operation and in 
emergency. The converter may be 
started by either of three general 
plans, viz. : a supply of current to 
either side of the machine or by ex- 
ternal power supplied by a starting 
motor direct-connected to the shaft. 
When started from the direct-current 
side, a rheostat is used in series with 
the armature, as in starting a direct- 
current motor. The starting current, 
however, has two paths, one through 
the converter windings from brush to 
brush, and another through the col- 
lector rings to the transformer coils 
and thence back again to the converter 
armature. While the converter is 
turning slowly the frequency of re- 
versal of current through the trans- 
former coils is low and the choking 
effect is small. The starting current 
from the direct-current side is, there- 
fore, more than that of a motor of the 
same size without load. When the 
machine has come up to speed the po- 
tential regulator is adjusted to bring 
the alternating-current pressure of the 
rotary converter up to that of the 
transmission system. The alternating- 

current side of the rotary is synchron- 
ized with the transmission system and 
connected to it. The field is adjusted 
to bring the power factor up to unity 
and the potential regulator is used to 
adjust the load carried by the unit to 
the desired amount. 

In case a total shut-down of the sys- 
tem removes the supply of direct cur- 
rent for starting, means must be at 
hand for starting from the alternating- 
current side with the field coils open 
as in starting a synchronous motor, 
and the pressure reduced to about half 
normal pressure to keep the starting 
current within limits. This may be 
done by means of a starting compensa- 

The starting current required in 
starting from the alternating-current 
side is from 150 to 200 per cent, of 
full-load current on a 500-kvv. con- 
verter and somewhat less on larger 
sizes. The direct-current starting 
current, however, is but 25 to 30 per 
cent, of full-load current, and this 
method is, therefore, preferred for 
regular operation. Sufficient machines 
are equipped for alternating current or 
motor on the shaft starting in a given 
substation to insure a supply of direct 
current for starting the other units. 
Where sufficient storage battery ca- 
pacity is installed the direct-current 
supply may be relied upon at all times. 

14 c Dia. 

Safety Stop-^ gsa 

Hearing ? |, 

[ W\\\gAmp: ' '' 





rrp inx 

ni 8"-X»a.- - 

Fig 10. 

tor on the high-tension side of the 
transformer, or by means of taps on 
the secondary winding. The latter is 
preferable, as no auto-transformer or 
extra high-tension switching opera- 
tions are required. 

In this method after the machine 
is brought up to speed, its fields are 
excited and the polarity noted, as it 
may come up reversed. If so, the di- 
rect-current voltmeter on the machine 
gives a negative reading. The field 
connections are then reversed by 
means of a switch provided for the 
purpose and the machine slips back 
one pole. As soon as it has done so 
the direct-current voltmeter swings to 
a positive reading, when the field is 
again reversed and the polarity re- 
mains correct. The starting switch is 
then thrown to the full pressure and 
the machine is equalized and thrown 
on to the direct-current bus. 

The smaller starting current re- 
quired in starting from the direct-cur- 
rent side makes this method jpreferable 
in cases where there are several ma- 
chines, or where the direct-current dis- 
tributing system has sufficient capac- 
ity to furnish the starting current 
without serious disturbance. In such 
cases the normal method of starting is 
from the direct-current end. 

The synchronous converter has also 
been adapted to operation on a verti- 
cal shaft after the manner of the fre- 
quency changer described in the fore- 
going. This machine is. however, 
supported on a bearing which operates 
on a pedestal that passes through the 
center of the machine to the top. The 
bearing is thus accessible from the top 
by the removal of a plate, instead of 
from below. The general arrange- 
ment is illustrated in cross-section in 
Fig. 10 and the external appearance 
in Fig. II. 



July, 1908 

These machines have been made in 
units of iooo and 2000 kw., the first 
of this type having been installed in 
Chicago in 1907. 

The use of motor-generators for 
conversion to direct current is some- 
times resorted to in preference to syn- 
chronous converters. When the trans- 
mission system operates at 60 cycles, 
the use of converters is subject to 
some disadvantages, and it is, there- 
fore, usual to find motor-generators 
in direct-current substations which de- 
rive their energy from a 60-cycle sys- 
tem. The inherent characteristics of a 

One of the principal advantages of 
the direct-current system of distribu- 
tion is the possibility of the use of a 
storage battery reserve. Before the 
use of the battery became general it 
was not an uncommon thing in the 
larger systems to have the service seri- 
ously interrupted through a compara- 
tively small accident in the generating 
or transmission system. With the in- 
troduction of the storage battery 
these interruptions were limited en- 
tirely to serious accidents affecting 
the major part of the equipment. 
Such smaller disturbances now occur 

Fig 11. 

60-cycle converter are such that it is 
very sensitive to fluctuations in fre- 
quency or voltage, which causes hunt- 
ing and sparking at the brushes when 
conditions are unfavorable. Such 
converters must be provided with 
copper rings about the pole pieces in 
order to damp the tendency to hunt. 
These difficulties are reduced in a sys- 
tem which derives its energy from 
turbine-driven generators, owing to 
the absence of a reciprocating motion 
in the prime mover. 

The motor-generator is subject to 
the handicap of greater first cost and 
lower efficiency than the converter. 
If induction motors are used the sta- 
bility of the system is increased in 
times of disturbances of short dura- 
tion, as induction motors will continue 
to operate when synchronous motors 
and converters are thrown out of step 
and shut-down. The necessarily low 
power factor of the induction motor 
may be partially compensated for by 
the use of both induction and syn- 
chronous motors in the same substa- 
tion, the synchronous motors being 
so excited as to supply the lagging 
current for the induction motors. 

in a large system without appreciably 
affecting the service. The usual ar- 
rangement of battery connections is 
shown in Fig. 12. The cells from 
which taps are brought out are known 
as end cells and are used as follows : 
Connection is made from each bat- 
tery terminal to a bus-bar by a slid- 
ing contact C, which bridges the gap 
between the bus-bars and the terminal 
as it is moved along. The voltage 
of each cell being about two volts 
the pressure delivered by the bat- 
tery to the bus-bar will vary accord- 
ing to the position of the sliding 
contact. When the battery is required 
to discharge the sliding contacts are 
moved toward the outer ends, thus 
raising the pressure of the battery 
and causing it to deliver energy to 
the bus-bar. When no energy is re- 
quired from the battery, the end-cell 
contact is set so that the battery pres- 
sure and bus pressure balance, and 
the battery floats on the system. In 
case of a reduction in the bus pressure, 
due to a failure in the supply of en- 
ergy, the battery pressure causes the 
battery to discharge to the bus, thus 
holding up the pressure partly and 

preventing a complete interruption. 
The extent of the interference depends 
upon the relative capacity of the bat- 
tery and the load on the bus at the 
time. During the hours of light load 
the operator's adjustment of the end- 
cell switches is sufficient to restore 
the pressure to normal in a very short 
time, so that the consumer notices 
nothing beyond a slight flickering in 
the lights. 

As the peak of the load in a large 
system is usually considerably greater 
than the average load, it is not feasible 
to provide sufficient battery to care 
for a serious accident at that hour. 
The chances of the break-down oc- 
curring at this time being rather re- 
mote, and the maintenance of bat- 
teries being expensive, it is not usual 
to provide more than 25 to 40 per 
cent, of the maximum load in battery 

Two buses are provided, so that the 
battery may discharge simultaneously 
to main and auxiliary buses at dif- 
ferent pressure. It is also desirable 
to keep the battery floating on the 
main bus while it is being charged 
through another bus. This battery 
may be charged through a booster 
from the main bus, or from a separate 
converter or generator wound for the 
higher pressure required for full 

The battery as installed in Ameri- 
can practice is usually arranged for 
motor control of the end-cell switches 
with indicators on the switchboard to 
show the operator the position of the 
end-cell switches on each bus, am- 
meters on each bus reading both ways, 
and pressure connections by which the 
voltage of each end cell may be as- 
certained whenever desired. 

The method of operating, and main- 
taining a battery involves a great 
many important details, which need 
not be elaborated in this connection, as 
they are fully covered in special works 
covering all phases of the subject. 

Circumstances make it advisable in 
some cases to combine direct current 
with an alternating-current substation, 
the direct current being distributed in 
the immediate vicinity, and the alter- 
nating current being used for a scat- 
tered load beyond the range of the 
direct-current lines. 



"h l 



Fig 12. 



DICKENS, in his usual extrava- 
gant way, in "Prince Bull" says 
this: "She was a fairy, this 
Tape — she could stop the fastest thing 
in the world, change the strongest into 
the weakest, and the most useful into 
the most useless. To do this she had 
only to put her cold hand upon it and 
repeat her own name, Tape. Then it 
would wither away."* 

Now we are told that hard in the 
wake of combinations, consolidations 
and nine-hundred-and-ninety-nine- 
year leases there sneaks this same bad 
fairy, Tape. But — do you believe in 
fairies ? I say, don't ;at least not in busi- 
ness fairies ; it is a poor policy. Deal 
with facts absolutely and entirely, and 
fairies, good or bad, will never trouble 
you. They are created merely for the 
weak, who try to hide their own short- 
comings behind the skirts of fancied 
grievances. Never mind what Dickens 
says — tape is a necessity. With the 
growth of every business the time 
comes when it begins to be imprac- 
ticable to have Tom, Dick and Harry 
meddle in everything and anything 
that concerns the company. The man 
who cannot bring himself to realize 
that had better quit right then and 
there, for tape — red tape, if you will 
— there must be to prevent chaos ; but 
those charged with its manufacture 
face a grave responsibility. 

Law is imperative, yet "that coun- 
try is best governed which has the 
least law." Rules must be designed 
to assist individual judgment, not to 
stifle it. The attempt to cover every 
specific instance by a general rule 
promotes indifference. Write all the 
rules you please while the spirit moves 
you, but never send them out till you 
are sober. There is the possibility of 
making the remedy worse than the 
evil. Remember the Shildburgers ? 
They complained to the town council 
that the game-warden on his rounds 
trampled down too much of their 
grain. "We will soon remedy that," 
decreed the council, and forthwith ap- 
pointed four strong men to carry the 
warden. Don't emulate that example. 

What is needed most is a con- 
scientious cultivation of the sense of 
proportion. For instance : I claim 
that when my boy was taught in school 
that Africa has 174,396,413 inhabi- 
tants a great wrong was done him. 
Such perverted accuracy must unduly 
twist the sober sense of judgment of 

*Read before the National Electric Light Asso- 
ciation at its Thirty-first Convention, held at 
Chicago, III., May 19-22, 1908. 

the relative importance of things. 
How many of us keep on counting the 
inhabitants of Africa to the last man 
all through our lives? Did you ever 
see a gang of forty-cent-an-hour line- 
men wait patiently — and willingly, for- 
sooth — while a nine-dollar-a-week 
storeroom clerk figured out the cost 
of a four-inch spike to the seventh 
decimal point of a cent? It is a sign 
of greatness to be able to judge when 
the seventh decimal is worth while. 
Perhaps it is most essential in scien- 
tific investigation, for behind it Nature 
may hide one of her well-guarded 
secrets, but the seventh decimal in 
every-day routine is an unmitigated 
nuisance. Yet, says the auditor, we 
must have it to keep our pennies 
straight. True, a penny one way or 
the other may be worth a war, if it 
involves a principle, but the man of 
the hour is he who can find the way 
to keep principles and pennies at a 
safe distance. If you insist upon 
carrying the supposed accuracy of 
your accounting one single point be- 
yond that warranted by the limit of 
correctness of its basic data you are 
foolishly and recklessly wasting time, 
money, and the very life-blood of 

Intelligent approximation is a high 
art sorely neglected. Figures should 
not be relied upon to the exclusion of 
common sense. Of course figures 
don't lie, but for that very reason they 
furnish the liar his sharpest tools. We 
are apt to become slaves of figures 
and worship the golden calf of aver- 
ages and percentages. At the end of 
every month we reduce an avalanche 
of reports to a few neat sheets of 
bare figures, a sort of skeleton de luxe, 
the dry bones of which are rattled in- 
dustriously in executive session and — 
well, the devil may catch the hindmost, 
or vice versa. Chasing figures with 
explanations is a sad job. Mere 
figures climb to places where explana- 
tions cannot follow. So much for 

Furthermore, in laying down the 
laws for the company the intimate in- 
terdependence of the departments 
must have most careful consideration, 
else there can be no successful parallel 
operation. The necessity for special- 
ization under modern conditions often 
places at the head of a department a 
man whose knowledge of the opera- 
tion of other departments is confined 
to what might be termed a course in 
the intra-company correspondence 
school, and in that case the value of 

an efficient clearing-house cannot be 
overestimated. Without it rules will 
clash, the same information will be 
asked for in half a dozen different 
ways, multiplying uselessly the labor 
necessary to get it together, the gen- 
eral good will be sacrificed to the 
whim of the individual, meritorious 
effort will be strangled, ambition em- 
balmed, and a dead "what's the use?" 
feeling will begin to permeate the sys- 
tem. The conclusions of a man who 
refuses to look beyond his own nose, 
no matter how long that may be, will 
lead him into the wilderness. 

Still another point. Nothing will 
promote economy more effectually 
than a continuous uniform rate of pro- 
duction. We fully realize that in the 
operating department, and are moving 
heaven and earth, and justly so, to 
find means to flatten the peak unalter- 
ably imposed upon our stations by the 
revolution of the earth once in twenty- 
four hours; yet we deliberately create 
an entirely artificial monthly peak in 
our commercial, accounting and busi- 
ness department, for which, in the last 
analysis, no more valid reason can be 
advanced than that the moon travels 
around the earth somewhere in the 
neighborhood of once a month. What 
a boon the elimination of this peak 
would confer upon an army of weary 
men, whose energies are indeed fully 
taxed in handling the tape that most 
careful and wise administration de- 
mands ! 

In handling this tape there are really 
but two points to be observed : they are 
strict compliance with the rules and 
the avoidance of errors. An immense 
amount of smooth red tape can be 
reeled off in record time, provided you 
take care not to get it into a snarl. 
In the complex system of a large con- 
cern errors are positively hydra- 
headed, and not every clerk is a 
Hercules. The only way to treat an 
error is to kill it a-borning; once it 
sees the light of day it devours time 
and paper in prodigious and ever-in- 
creasing quantities. A runaway error 
is harder to stop than a slander. 

With all due precautions taken to 
avoid them, there will crop up now and 
then, among the complexities of neces- 
sary tape, some really weired ex- 
crescences. How shall we treat them ? 
Little good will result from combat- 
ing ingenious foolishness with till— 
manesque ferocity ; sarcasm will al- 
ways be paid for in counterfeit coin. 
Neither is it well to try to do a rule 
to death by quixotic superobservance ; 




July, 1908 

you run the risk of being done first. 
Nor yet is it well to skip over it in a 
careless, perfunctory way, like that 
particular type of Bridget who is al- 
ways done. According to her notion, 
dinner was invented for the sole pur- 
pose of having it over with. It is the 
wiser plan to use sober reason per- 
sistently ; nothing inherently wrong 
can stand long before it. 

But there remains one case where 
individual judgment must rise above 
all rules — that is, an emergency. An 
emergency, provided he has taken 
every precaution to prevent it, is a 
man's greatest opportunity. He who 
in the fierce glare of an emergency 
hides in the protecting shadow of a 
rule might as well acknowledge him- 
self a coward, for, if he is a man at all, 
he will forever after carry with him 
the gnawing consciousness of failure. 

After all is said, the backbone of a 
company is not its rules, but its men. 
No concern of magnitude has any use 
for heroic individualism of the hermit 
type, but there is a crying need for 
honest, loyal, collective individualism. 
Out of the maze of filing cabinets, 
where card indices innumerable gather 
dust in their polished metal-trimmed 
caskets ; above the bevies of typists 
who clatter away feverishly at the 
keys between primping and candying ; 
above the horde of clerks wallowing 
in quadruplicates ; above the groups of 
petty officials, who deem it their prin- 
cipal duty to dictate reams of letters 
to be handed across the corridor — 
there must always rise a select few, 
firm of grip, clear of view, broad of 
mind, who will absolutely dominate 
red tape. Try, you, to be one of them. 
Relegate relentlessly to some one else 
all the things that some one else can 
do as well as yourself. Thoroughness 
does not mean indiscriminate attention 
to an agglomeration of petty details ; 
it means the intelligent elimination of 
unessentials and -a firm grasp on mat- 
ters of vital importance. Let it be 
your constant aim so to manage affairs 
in your care that they will run 
smoothly and efficiently without you ; 
for the moment you have reached that 
point invariably coincides with that of 
your promotion. The battle will al- 
ways be to the strong, sacred as well 
as demagogic generalities to the con- 
trary notwithstanding. Never flag. 
Retrogression begins the moment you 
are firmly convinced that you are 
doing your best ; there really is no best. 
Hundred per cent, efficiency and per- 
petual motion are synonymous. Meri- 
torious effort and genius forever try 
to approach them ; ignorance and mad- 
ness alone claim to reach them. Be 
always on guard, for as soon as you 
begin to count upon what you have 
done yesterday, and not what you are 
doing to-day or what you are able and 

willing to do to-morrow, it is the 
proper time to hand in your resigna- 
tion if you would avoid unpleasant- 
ness. When you arrive at that state of 
mind where everything anybody else 
does is absolutely no good ; when all 
changes and innovations are utter rot ; 
when you feel more like throwing a 
handful of sand into the business ma- 
chinery than using the oil can ; when 
you become a slave of conditions in- 
stead of being their master, you are 
old, brother — you are ripe for the pen- 
sion list and half pay. 

The company man needed and 
wanted is he who can stoutly and 
with ability defend his own opinion 
while a matter is under discussion and 
open to argument, and who, if over- 
ruled, is broad enough to bring his 
best efforts, with true loyalty, to bear 
upon making an unqualified success of 
the opinion that prevailed ; in short, 
a company man must be able to abide 
by the decision of the court and go to 
work. Concerted action of men who 
have minds of their own must always 
be based on a compromise. Though 
we cannot all get what we think is 
best, we must nevertheless all do the 
best we can with what we get. He 
who goes off nursing a grudge, sulking 
behind petty routine, waiting and 
watching for the time when he may 
cra~wl out of his hole with a doleful 
"I told you so," will soon find himself 
forgotten. The chances are that he 
will be shriveled and dry before his 
opportunity arrives. 

But in all our efforts let us never 
forget that, besides being company 
men, we are also men — just that. Let 
us not fail to observe that the world 
is slowly recasting valuations. Raw 
dollars are not worth as much as they 
were some years ago, still there are 
men who, having sold their souls to 
success, keep on counting dollars 
above all else, and they will probably 
die doing it. My appeal is to the 
young men — you whose actions will 
determine the fate of the nation for 
the next generation — yet I cannot find 
words adequate to take the place of 
these, which, though spoken over 60 
years ago, seem more timely to-day 
than ever before. I will but humbly 
repeat them in the fond hope that you 
may heed their lesson : 

" — And, further, I will not dis- 
semble my hope that each person 
whom I address has felt his own call 
to cast aside all evil customs, timidities 
and limitations, and to be in his own 
place a free and helpful man, a re- 
former, a benefactor, not content to 
slip along through this world like a 
footman or a spy, escaping by his 
nimbleness and apologies as many 
knockes as he can, but a brave and 
upright man, who must find or cut a 
straight road to everything excellent 

in the earth, and not only go honorably 
himself, but make it easier for all who 
follow him to go in honor and with 

"After all is said, the backbone of 
a company is not its rules, but its 

Something' WortH Knowing About 
Direct-Current Meters 

Tests on a certain type of D. C. 
meter show that a bus bar carrying 
600 amperes at a distance of 12 inches 
from the meter was found to affect its 
accuracy at one-tenth load over 50 per 
cent., the direction of the current in 
the bus bar determining whether the 
meter was this percentage fast or slow. 
The error is largely dependent on 
the position of the bus bar with rela- 
tion to the meter, as the same current 
with the bus bar only two inches from 
the meter in another direction had no 
appreciable effect on its accurary. In 
actual practice a case is reported of a 
600 ampere D. C. meter in perfect 
condition which would not register on 
a load of 80 amperes on account of the 
opposing field from an adjacent con- 
ductor. Meters may also be influenced 
by their own wiring, if the service 
or load wires are brought around the 
meter. In another case reported from 
actual practice, a 150 ampere D. C. 
meter registered only 90 per cent, of 
the load passing through it on this 

Meters are also somewhat affected 
by their proximity to each other. In 
tests on D. C. meters where three 
meters were placed side by side with 
12 inches between centers, it was 
found that the middle meter was 
affected from five to 10 per cent, on 
one-twentieth load when the other two 
carried a full load. To be free from 
this effect meters should be installed 
with about 15 inches between centers. 
While the instances cited are from 
D. C. meters, some makes of A. C. 
meters are nearly as much affected by 
external fields as are the D. C. meters. 

It will at once be seen that the 
installation of meters on switchboards 
or • close together in meter closets, 
where they are also frequently very 
near the building risers, may give rise 
to considerable metering errors which 
might be avoided by a more judicious 
location of the meters. The matter is 
one which merits the careful attention 
of our companies and information can 
doubtless be obtained from the meter 
manufacturers if requested regarding 
possible errors from this cause. There 
has recently been designed for use 
where external fields cannot be 
avoided, a four-pole D. C. meter which 
is much less affected than the ordinary 
two-pole type. — Meter Committee Re- 
port — A'. E. L. A., 1908. 

Receiving Stations Operated from High-Tension 

Transmission Lines 


THE design of receiving stations 
operated from high-t e n s i o n 
transmission lines has become of 
vital importance to those responsible 
for obtaining the best results, and the 
main features involved in such station 
design warrant careful consideration. 

The term '"receiving station," as 
used in this paper, covers stations 
where practically all of the electrical 
power is received over one or more 
incoming feeder circuits from other 
stations where it is generated by steam 
or water-power. Auxiliary steam or 
water-power may be available in a 
receiving station for an emergency or 
to assist in carrying the peak load, 
but the bulk of the power is received. 

In taking up the details of the ap- 
paratus used in such stations, it will 
be obviously impossible even to touch 
on all of the different types, so one 
particular line of apparatus will be 
considered and compared with other 
conflicting designs.* 

The matter of suitable equipment 
for receiving stations will be consid- 
ered under the following heads : 
(i) General Features 

(a) Type of Station 

(b) Type of Transformer 

(c) Main Connections 

(2) Necessary Equipment 

(a) Switchboards 

(b) Circuit-Breakers 

(c) Disconnecting Switches 

(d) Protective Devices 

(e) Bus Bars and Wiring 

(f) Auxiliary Apparatus 

(3) Present Design 

(a) Switching Stations without 


(b) Receiving Stations with 

Type of Station — This paper deals 
with receiving stations operated from 
transmission lines of 22,000 volts and 
upward, with their equipment of 
transformers, switching devices, pro- 
tective apparatus, and so forth. To 
secure the most suitable arrangement 
of a receiving station it is essential 
that the building be designed for the 
apparatus that it will contain, instead 
of attempting to arrange the equip- 
ment in a building already erected. 
It is the purpose of this paper to point 
out the main features of the apparatus 
required in receiving stations, with 
particular reference to such points as 
influence the design of the building. 

*Xational Electric Light Association. 1908. 

Type of Transformers — Trans- 
formers may be divided into types 
determined by the means adopted for 
dissipating the heat developed in 
operation. These means may be di- 
vided into natural air-cooling, artifi- 
cial air-cooling, natural oil-cooling, 
and artificial oil-cooling, and these de- 
termine the main features of the 

Natural air-cooling depends on the 
conduction of heat through the air 
from the transformer to its case, 
where it is radiated to the surrounding 
air, and it has been found imprac- 
ticable to build such transformers in 
large sizes. 

Artificial air-cooling is obtained by 
forcing a blast of air through the iron 
and the coils of a transformer placed 
in a cast-iron housing, and ordinarily 
located over an air chamber where a 
pressure of one-half to one and one- 
half ounces is maintained by a blower, 
usually driven by a motor. 

Owing to the difficulty of securing 
satisfactory insulation without the use 
of oil, air-blast transformers are not 
built for pressures above 33,000 volts, 
and are not recommended for more 
than 22,000 volts. 

Natural oil-cooling takes place 
when a transformer is immersed in 
oil and provided with a case having 
sufficient radiating surface to dissipate 
the heat brought to it by the oil. 
Transformers of this type have nu- 
merous ducts extending through the 
opening in the iron from one end of 
the coil to the other, and similar 
ducts between the laminations permit 
a circulation of oil throughout the in- 
terior of the transformer. 

Artificial oil-cooling is adopted for 
larger sizes, and two methods are in 
general use for cooling the oil, these 
being known as forced water and 
forced-oil circulation. In the former 
method one or more coils of seamless 
brass tubing are placed under the oil 
in the top of the transformer case 
near the walls of the tank, and water 
is circulated through these coils to 
remove the heat from the oil and to 
promote the circulation of the oil. 

The circulation of oil in such trans- 
formers depends on the difference in 
weight between the hot and cold oil. 
and where the transformers are of 
large capacity this natural circulation 
is not vigorous enough to carry off the 

heat developed, and it is necessary to 
resort to artificial circulation. This 
is accomplished by pumping the oil 
from the transformer case and circu- 
lating it through cooling coils placed 
in running water, a surface condenser 
or a blast of air. 

For units of 5000 kilowatts and 
above, forced-oil circulation becomes 
advisable as giving more uniform 
cooling, lower maximum and lower 
average temperature rise. Three-phase 
transformers of 10,000 kilowatts for 
1 20,000- volt service have been de- 
signed with "forced-oil circulation, and 
even larger units can be built if re- 
quired. The limit of size is largely a 
question of shipping facilities or of 
building up the transformer at desti- 

The question of three-phase or 
single-phase transformers has been 
ably summed up by Mr. J. S. Peck in 
stating that three-phase transformers 
when compared with three single-phase 
units of the same total capacity, have 
the advantage of lower cost, higher 
efficiency, smaller floor space, less 
weight, simpler wiring, reduced freight 
and erection charges ; while the main 
disadvantages are the greater cost of 
spare units and repairs. Allowing a 
spare unit in each case for one or two 
circuits, the single-phase units have 
the advantage in price of transform- 
ers, while for three or more circuits 
the three-phase units have the ad- 

The relative advantages of star and 
delta connections have been pretty 
thoroughly discussed at various times 
and seem to be reduced to these fea- 
tures. With delta connections one set 
of coils on a three-phase transformer 
or one of a bank of three single-phase 
transformers can be cut out of circuit, 
and the remaining two can then be 
connected in open delta and deliver 
173/300 of the output with the same 
heating. With star connections the 
voltage on any one transformer is only 
0.577, tne amount between lines and a 
neutral point is available for ground- 
ing or other purposes. 

The question of using star connec- 
tions and a neutral grounded solidly 
or through a resistance is a mooted 
one, but the general practice, particu- 
larly for very high voltages, favors 
the star connection with neutraf 
grounded through a resistance that 




July, 1908 

will limit the current that flows 
through a grounded line. 

The use of small transformers oper- 
ating from extremely high-tension 
lines is complicated by the fact that 
a power transformer cannot be eco- 
nomically built for a smaller capacity 
than about 0.5 kilowatt per thousand 
volts. As the voltage of the lines is 
raised, the size of transformer that 
can be economically built for opera- 
tion at that voltage goes up, and it 
occasionally happens that a small con- 
sumer located near a transmission line 
of 66,000 volts, or higher, desires an 
amount of power less than that for 
which a transformer can be furnished 
suitable for that voltage. 

On some of the larger transmission 
circuits, such as those of the Niagara, 
Lockport and Ontario Company's 
lines in New York, transformer sta- 
tions are installed for lowering the 
line pressure from 60,000 to 11,000 
volts, and various customers are then 
supplied at 11,000 volts. 

Main Connections — For switching 
stations located along a transmission 
line the main connections are very 
simple, and, as a rule, simply allow 
for the opening up or connecting to- 
gether of the various lines. For re- 
ceiving stations with transformers the 
main connections will depend on the 
number and capacity of incoming 
lines, step-down transformers, outgo- 
ing ieeders, local circuits, and so 
forth, and the proper arrangement of 
the bus bars, and so forth, is a matter 
of great importance. With one in- 
coming line and a transformer circuit 
of the same capacity as the line, no 
high-tension bus bars are needed, as 
the line will connect through suitable 
switching devices direct to the trans- 
former circuit. 

If the one line feeds two or more 
transformer circuits, or two or more 
lines feed one transformer circuit, a 
single set of high-tension bus bars will 
be required. If there are two or more 
lines and two or more transformer 
circuits, a single set of bus bars is 

With three or more lines, the best 
results can usually be obtained by hav- 
ing the transformer circuits of the 
same number and capacity as the lines, 
and arranging so that, normally, each 
line will feed its own transformer cir- 
cuit, but in case of necessity can be 


s< tmetimes used ; but. as trouble in this 
bus involves a complete shut-down of 
the station, a more flexible system is 
advisable. This flexibility is obtained 
by using a sectionalized bus, double 
bus, ring bus or relay bus, depending 
on the number of lines and circuits 
and the amount of flexibility desired. 

connected to any transformer circuit. 
Switchboards — In some of the 
switching and transformer stations 
described under division 3, various 
knife-switches, circuits, breakers, and 
so forth, are provided, and no switch- 
board, properly so called, is used. In 
other stations, particularly where elec- 

f Chonne/ 1, 

Cena.J pf fm MM *» ce-*-a*^n pant! 


FRON'* I /.\ 

£so {t(v*ru\ 


• /-■ 


C»nt-#ltt /> fimnS C'Smf 
'- /«•« 
4.1 >§tt > #•>• 

Cn*ffi$tttrs far ftt&f C- gtt,% 

Pi AN V'StV 


July, 1908 



trically operated, oil circuit-breakers 
are used, some central point is usually 
chosen from which to control the sta- 
tion. This central point is ordinarily 
a switchboard on which are mounted 
the controlling devices, relays, and so 
forth, as well as various meters for 
the transformer and feeder circuits. 
These switchboards for receiving sta- 
tions are usually made of the panel 
type, although occasionally a control- 
desk or bench-board is supplied. 

The switchboard panels or the top 
slabs of the control-desk are ordinarily 
made of polished marble or slate with 
oil or marine finish. While polished 
marble has been used in many instal- 
lations, marine-fish slate is growing 
in favor more and more, owing to its 


being somewhat cheaper, and far 
easier to match and to keep in first- 
class condition. 

The panel-boards used for receiving 
stations are simply modifications of 
standard switchboards, with panels 
for various line, ' transformer and 
feeder circuits, as well as a panel for 
the control of a small storage battery 
with the motor-generator set used for 
charging it. Fig. i shows a switch- 
board of this type supplied to the Ni- 
agara, Lockport and Ontario Com- 
pany for use in its Gardenville sub- 
station for supplving power near 

Where there are a large number of 
circuits to be controlled and a very 
compact arrangement is desired, a 
control-desk is often used, with the 
controllers mounted on the horizontal 
top of the desk and the instruments 
back of and above the desk in such 
a manner that the operator at the desk 
will face the switching and trans- 
former room. While operating the 
controllers, and so forth, he can watch 
the meters or look under the instru- 
ment frame and over the top of the 
desk and watch the breakers, and so 

Fig. 2 shows a combined desk 
and instrument frame of this type, to 
be used ultimately to control five 9000- 
kw., 60,000-volt, three-phase incoming 
lines; five 9000-kw., 60,000-12,000- 
volt, three-phase step-down transform- 
ers, one 1 2,000- volt local service feeder 
and 29 12,000-volt distributing feeders. 
The original installation requires two 
of the five sections and will control 
two lines, two transformers and 12 
distributing and one local service 

The general connections of this 
station are rather clearly shown by 
means of the miniature bus-bar system 
on the top of the desk. The general 
scheme is to normally have each in- 
coming transmission line feed its own 
three-phase transformer, and each 
transformer will normally operate on 
its own auxiliary bus, supplying power 
to about six 12,000-volt feeders. 

The top of this control-desk consists 
of marine-finished slate slabs, while 
the front, back and sides are made of 
steel plates that are readily removable 
to permit access to the interior of the 
desk for getting at the connections, 
adjusting the controllers, and so forth. 
Circuit-Breakers — While oil cir- 
cuit-breakers are ordinarily employed 
to furnish automatic protection in 
high-voltage circuits, fused breakers 
of the type shown in Fig. 3 have been 
used on the circuits of the Niagara, 
Lockport and Ontario Power Com- 
pany, to cut off the individual trans- 
formers, and fused breakers of this 
same general design have often been 
used in substations for single-phase 
railway propositions, as shown in 
Fig. 11. 

Fig. 3 shows clearly the general ar- 
rangement of a fused circuit-breaker 
mounted on line insulators that are 
provided, suitable for the line voltage, 
and the design is such that the breaker 
arm can be arranged to open parallel 
to or at right angles with the plane 
of the support to which the base is 

This fused circuit-breaker is essen- 
tially a single-pole device, and for 
special applications, such as cutting 
off individual transformers in the 

manner used by the Niagara, Lock- 
port and Ontario Power Company, or 
in transformer stations of single-phase 
railways, they have given great Satis- 
faction. For the three-phase work it 
is usually advisable to have all three 
of the main circuits opened at the 
same time, and oil switches and cir- 
cuit-breakers are used for this pur- 

The essential feature of an oil 
switch or breaker is that the circuit 
is ruptured under oil, and designs 
have been perfected for all classes of 
service, from the small-capacity, low- 
voltage apparatus to the 60,000-volt 
breakers used on the circuits of the 
Ontario Power Company, which are 
guaranteed to operate satisfactorily 
under any condition of overload or 
short-circuit that might exist in a plant 
of 200,000-kw. capacity. Breakers 
have been built and designed for cir- 
cuits up to 132,000 volts, and higher 
limits can be reached if necessary. 

Owing to the amount of power re- 
quired for operating large oil circuit- 
breakers, motors or solenoids are usu- 
ally employed, although practically 


any type of solenoid-operated breaker 
can be arranged for hand operation. 

Fig. 4 shows one pole of a breaker 
that is built in capacities up to 1200 
amperes at 3500 volts and 100 am- 
peres at 33,000 volts, with an ultimate 
breaking capacity for three units, 
forming a three-pole breaker of 10,400 
kw. Each pole is intended to be 
placed in a separate fireproof com- 
partment of masonry and is provided 
with its own closing and tripping 
solenoid, these being operated in mul- 
tiple by a single controller. The de- 
sign of the mechanism, contacts, and 
so forth, is clearly shown. 



July, 1908 


is shown directly above the closing 
solenoid, but where electrical opera- 
tion is not desired it is possible to 
place the handle on the switchboard 
and operate the breaker by means of 
a suitable bell-crank mechanism. 

Fig. 7 shows a breaker built for 
60,000-volt service and guaranteed to 
operate satisfactorily under any condi- 
tions of overload or short-circuit that 
might exist in a plant having 200,000 
kw. in generating capacity. Modifi- 
cations of this breaker can be arranged 
for high voltages. 

The breakers shown in Figs. 6 and 
7 are essentially top-connected, self- 
contained solenoid breakers with metal 
tanks, and these features are particu- 
larly valuable for the class of service 
for which they are intended. 

The top-connected breaker is made 
with metal tank, and there is no 
trouble in securing oil-tight joints. 

Fig. 5 shows a group of breakers 
that are built in capacities up to 3000 
amperes at 3300 volts, and 300 am- 
peres at 35,000 volts, and having an 
ultimate breaking capacity that has 
never been reached in any plant now 
installed or contemplated. Each pole 
of this breaker is enclosed in masonry 
structure, and all of the poles are 
operated by a single powerful mechan- 

The breakers shown in Figs. 4 and 
5 are primarily designed for use in 
plants having the cellular construc- 
tion for bus bars, wiring, and so forth, 
where it has been found of the utmost 
importance ,to .isolate ,the ,bus ,bars, 
wiring, and so forth, in such a man- 
ner that leads of opposite polarity are 
separated by soapstone, concrete, 
brick or similar material, to prevent 
an arc starting in one place from be- 
ing communicated to an adjacent con- 

The amount of current available 
momentarily at the point of trouble 
in large stations operating at pres- 
sures of 12,000 volts or less is some- 
thing enormous, and every precaution 
must be taken to prevent the spread 
of trouble. The problem of suitable 
distances and insulation is a simple 
one for such voltages. 

For the high-tension circuits of 33,- 
000 volts and above, the question of 
enclosing the bus bars and wiring be- 
comes an entirely different proposi- 
tion, for the reasons given under (2)- 
(e), "Bus Bars and Wiring," and it 
is highly desirable to use an open sys- 
tem of wiring with breakers particu- 
larly designed for that class of work. 

Fig. 6 shows a breaker designed for 
60,000-volt service with an ultimate 
breaking capacity of 20,000 kw., three- 
phase, while a modification of this 
breaker has been built for 88,000-volt 


pacity of 40,000 kw., three-phase. The contacts are near the top of the 

The hand-closing device furnished tank, where the oil is apt to be in 

service with an ultimate breaking ca- with the electrically operated breaker better condition than at the bottom, 

Jaly, 1908 



and little trouble is found due to sedi- 
ment, and so forth, settling on the 

The top-connected, high-voltage 
breaker is provided with a single 

overload, reverse current or other 

The problem of suitable relays for 
use in connection with two or more 
incoming lines operating in multiple 


direct-pull solenoid, located near the is an extremely difficult one. An over- 
floor where the station attendant can load, ground or short-circuit on one 
readily inspect the mechanism, which line will draw current from all of the 
in case of trouble will tend to fall 
open rather than close. 

The top-connected breaker has all 
of the live metal parts submerged in 
oil, with the tanks, framework, mech- 
anism, and so forth, thoroughly 
grounded and the breaker entirely 
self-contained, while with breakers 
having the terminals at the bottom 
of the pots, and the plunger rods that 
go into the top of the tanks exposed, a 
masonry structure is necessary. 

The top-connected breaker can be 
built suitably for outdoor service in 
the manner shown in Fig. 8, which 
shows the outline and overall dimen- 
sions of a hand-operated breaker for 
use on a no,ooo-volt circuit for out- 
door service. Particular precautions 
have been taken for rendering the 
operating mechanism, terminal bush- 
ings, and so forth, impervious to se- 
vere weather conditions. 

In order to furnish automatic pro- 
tection to alternating-current circuits, 
relays of various types are used for 
closing the tripping circuits of the oil 
breakers. These relays can be made 
to operate instantaneously or with a 
time limit, either adjustable or in- 
verse, and to give protection against 

was to provide reverse-current relays 
on the line circuits at the receiving 
station, these being operated when a 
damaged line drew power from the 
bus instead of delivering power to it. 
As overload protection was also 
needed, it usually happened that the 
overload relays on the good lines 
would act as quickly as the reverse- 
current relays on the damaged line 
and all of the breakers would trip out. 
This was remedied by providing time 
limit for the overload relays and mak- 
ing the reverse feature instantaneous. 

Another trouble arises from the 
fact that in case of a serious short- 
circuit near the end of a transmission 
line, the heavy current flowing reduces 
the voltage and power-factor at the 
receiving station to such a low value 
that the reverse relay, whether built 
on the differential principle or the 
wattmeter principle, will not develop 
enough torque to act. A modification 
of the wattmeter type of relay has 
been built to give practically the same 
torque at low power-factor as at high, 
and to operate as a current relay even 
when the voltage drops to zero. In- 
verse time element features have been 
introduced to exercise a selective influ- 
ence and to trip out the circuit in 
trouble without interrupting the other 

The wattmeter type of relay, if set 
to a sufficiently delicate point to act 
on low voltage with sufficient torque 
on the contacts, is liable to trip out, 
due to vibration or mechanical shocks 
or to instantaneous surges, and no in- 
verse time element feature can be 




the receiving-station 
bus bars, and is apt to trip out all of 
the breakers and shut down the entire 
station unless relays are provided that 
will cut out the damaged line without 
affecting the others. 

The first solution of this problem 

worked in. The contacts have the 
duty of closing the trip circuit and, 
due to inherently low torque necessary 
to secure sensitiveness, they are liable 
to "freeze" and to have other mechan- 
ical troubles. A device known as the 
selective watt relay has been developed 



July, 1908 

for selecting 1 between reverse and 
overload. This relay is of the "nor- 
mally closed type" and is connected in 
series with the overload and reverse- 
current relay in the tripping circuit of 

seems to be necessary to obtain the ducting material getting across the 


the breaker. The smallest flow of en- 
ergy in the normal direction opens the 
contact of the selective watt relay, ren- 
dering the other relay inoperative. If 
a reversal equivalent to one per cent, 
occurs, the contact of the selective 
watt relay closes and the main relay 
can trip out the breaker. Owing to 
the time limit feature of the main 
relay, sudden surges will not trip out 
the breaker. 

If overload protection is also de- 
sired, this can also be obtained by 
having independent closing contacts 
on the wattmeter relay; those on the 
overload side connecting direct to the 
tripping circuit of the breaker, while 
those on the reverse side connect 
through the contacts of the selective 
watt relay. The wattmeter relay can. 
of course, be provided with time-limit 

Protective Devices — For high-ten- 
sion circuits, choke-coils and lightning 
arresters are provided, to furnish pro- 
tection against lightning and static 
disturbances of various kinds. 

Choke-coils for such circuits are 
usually made in the form of a .flat 
spiral for circuits up to about 25,000 
volts, while for high voltages either 
the oil-immersed type or the open 
helical type can be supplied. As a 
rule the strain developed across ad- 
jacent turns of a choke-coil in a high- 
tension circuit is so great that when 
the coil performs the function nor- 
mally expected of it, oil insulation 

best results 

Fig. 9 shows an oil-immersed choke- 
coil for use on an 88,000-volt trans- 
mission circuit, where the lightning 
conditions are particularly severe and 
where the maximum amount of pro- 
tection possible was desired, owing to 
the importance of the circuits to be 

Lightning arresters for these re- 
ceiving stations are made of the low- 
equivalent multi-gap or of the electro- 
lytic type, and as these have been often 
discussed and described there is no 
necessity of taking them up in this 
place. The electrolytic type of ar- 
rester seems to be growing in favor, 
largely due to the fact that it is suit- 
able for outdoor service. 

Bus Bars and Wiring — Practically 
all large receiving stations enclose the 
low tension secondary wiring if, be- 
tween the limits of 500 and 13,200 
volts, owing to the enormous momen- 
tary amount of current available under 

bus bars. 

For voltages of 33,000 and above 
fireproof barriers and cellular con- 


/*»r«sr*f/ry f*r>* 


short-circuit conditions with the disas- struction are unnecessarv, as the vio- 

trous arcs resulting. This provides lence of the arc and the destructive 

protection to the operator and security effects of a short-circuit depend on the 

against breakdown due to some con- amount of current available at the 

July, 1908 




.'13 ixl.: __n__L' _vl.lil...:_;j .".civ'uvti cx- 



"I — 



3f-C — 

- Jt -o- 


-+— 7*»^ 







J F i n - 'n H ri l' lR ilJii i Ldin 'i ir i' i i H i i l nnh h mn i i l n »» l » Jh I i i Li r un h I h ' u > ■ mn li -i lii ni u uu i 


Hir!»if«l r l?I^ir!UiP:ni«0JHi9*P?HjH»!J|niupft-'B 

Hi ! ! i i^TTi'rn r ' J ! I 

J 60 


July, 1908 

point of short-circuit, or for the same 
amount of power are inversely propor- 
tional to the voltage. 

Fireproof barriers offer a more or 
less perfect ground for high-voltage 

done only after power is cut off from tension and on the low-tension side, 

the circuit. Where branch lines run In the incoming-line circuit are two 

off from the main lines, fuses are fused type circuit-breakers to furnish 

sometimes used, these being made of the automatic protection, and two oil- 

a sufficient length of fine copper or insulated choke-coils with low-equiva- 

— — ■ " I 1 , -f\ 

_».„ — * — ^~r — """* ' 


circuits, and the higher the voltage the 
more perfect the ground. With bar- 
riers the striking distance to ground 
has been reduced 50 per cent, or more 
over what could be obtained with open 
wiring in the same space. 

Enclosed high-tension wiring, rec- 
ommended by certain engineers, re- 
quires a more expensive building and 
makes the inspection and repair of 
bus bars, wiring, disconnecting 
switches, and so forth, far more diffi- 
cult. This increased difficulty of in- 
spection prevents incipient trouble be- 
ing noticed so readily. 

Auxiliary Apparatus — In receiving 
stations such as form the subject of 
this paper the main circuits are all 
alternating, and as direct current is 
usually required for the oil circuit- 
breakers, and so forth, it is necessary 
to install a mercury rectifier or a 
direct-current generator driven by an 
alternating-current motor. Usually a 
small 125-volt storage battery of from 
40 to 100 ampere-hours' capacity is 
provided to insure a constant source 
of direct current, and this is charged 
by the rectifier or motor-driven gen- 

Switching Stations without Trans- 
formers — Such stations are used for 
sectionalizing transmission lines or for 
providing branch circuits from the 
main lines. As a rule, these consist 
simply of knife-type disconnecting 
switches, like Fig. 10, mounted on the 
towers or poles, and all switching is 

aluminum wire to insure the opening 
of the circuits. 

A typical example of station tapped 
in on a transmission line is shown in 
Fig. 11, which illustrates the general 
arrangement of a single-phase trans- 

lent lightning arresters and discon- 
necting switches. 

Fig. 12 shows the plan view and 
elevation of the top floor, as well as a 
section through the top and bottom 
floor, of the 44,000-volt receiving sta- 


former station for the Chicago, Lake 
Shore and South Bend Railway Com- 
pany. In this station there are three 
500-kw., 33,000-6600-volt, oil-insu- 
lated, self-cooling transformers, con- 
nected in multiple both on the high- 

tion of the Provincial Light. Heat 
and Power Company, of Montreal. 
This station takes care of two 44.000- 
volt incoming lines and two banks, 
each of three 2500-kw. step-down 
transformers, delta-connected high- 

July, J908 



tension and low-tension. Normally, 
each transmission line feeds its own 
bank of transformers, any one of 
which can be cut out of circuit by 
means of 44,000-volt disconnecting 
switches mounted on the wall. By 
means of the oil circuit-breakers in 
each incoming line and the tie-breaker, 
either or both lines can feed either or 
both banks of transformers. Each in- 
coming line is provided with three 
oil-immersed choke-coils and three 
low-equivalent lightning arresters 
with disconnecting switches. 

Fig. 13 shows the plan view, Fig. 14 
the elevation and Fig. 15 the section 
of terminal station that will ulti- 
mately control five 9000-kw., 60,000- 
volt incoming lines ; five 9000-kw., 
three-phase step-down transformers ; 
one 1 2,000- volt local-service feeder, 
and 29 1 2,000- volt distributing feed- 
ers. This is the station for which the 
control-desk shown in Fig. 2 will be 

used, and approximately two-fifths of 
the ultimate installation will be put 
in at first. The dotted lines appearing 
on Figs. 13 and 15 show the addi- 
tional space required for the installa- 
tion of three single-phase transform- 
ers in place of each three-phase trans- 

Fig. 13 shows the plan- view location 
of the apparatus in the low-tension 
and high-tension circuits. One part 
of the plan view shows the upper floor 
and locates the low-tension breakers, 
control-desk, terminal panels, and so 
forth; another portion shows the 
lower floor with the high-tension 
breakers, choke-coils, lightning ar- 
resters, and so forth. 

Fig. 14 shows the relative location 
of the step-down transformers, high- 
tension and low-tension circuit-break- 
ers, bus bars, disconnecting switches, 
and so forth. This elevation is taken 
in such a manner that one portion 

shows the 12,000- volt circuit-breakers, 
bus bars, connections, and so forth, 
while the other portions show the con- 
trol-desk, and so forth. One part of 
the high-tension compartment shows 
the oil circuit-breakers and step-down 
transformers, while another portion 
shows the out-going lines; choke-coils, 
lightning arresters, and so forth. 

Fig. 15 shows the sectional view of 
the station, locating the transformers, 
low-tension and high-tension oil cir- 
cuit-breakers, control-desk, relay pan- 
els, and so forth. As may be noted, 
all of the high-tension apparatus, bus 
bars, wiring, and so forth, have been 
located on the main floor, and can be 
readily inspected without danger, and 
without having to visit several floors 
and remove a large number of doors. 
The low-tension breakers, bus bars, 
and so forth, are placed on the second 
floor, with the control-desk, instru- 
ment and relay panels, and so forth. 

Distribution in Suburban Districts 


THE four-wire, three-phase system 
seems to have many advantages 
over other systems for suburban 
work.* In this system the generator or 
transformers are Y-connected, the neu- 
tral being brought out and connected to 
a grounded fourth bus bar. The voltage 
between any phase wire and neutral is 
2200 and between any two of the 
phase wires is 3800. Single-phase 
feeders are supplied from neutral and 
any phase being controlled by double- 
throw switches for balancing. Dis- 
tricts in which the load consists of 
both power and lighting are supplied 
from four-wire feeders, which run to 
centers of distribution from which 
radiate two-wire, single-phase mains 
for lighting, and three or four-wire, 
three-phase mains for power. If the 
load is mostly lighting the feeder may 
be three-wire, consisting of two-phase 
wires and neutral. In this case but one 
feeder regulator is needed, the extra 
phase wire being used for power only. 
Power in large units is supplied from 
three standard transformers, con- 
nected with primaries in star and sec- 
ondaries in delta. Small three-phase 
power installations may be supplied 
from two-phase wires and neutral by 
two standard transformers in open 
delta. This system requires but one- 
third the copper necessary for single- 
phase or four-wire, two-phase at 2200 
volts and about 44 per cent, of that 
required for a three-wire, three-phase 
system operating at 2200 volts under 
equivalent conditions. Mention has 
been made above of the necessity of 

♦National Electric Light Association, 1908. 

supplying as large an area as possible 
from each substation in order to keep 
down substation investment and main- 
tenance. The four-wire, three-phase 
system is a means to this end. Its use 
makes it possible to distribute over a 
much larger area than can be reached 
economically by either the two-phase 
system or the three-wire, three-phase 
system, and this advantage is secured 
without the necessity of departing 
from the use of standard transformers. 
A factory power installation of 200 or 
300 horse power at a distance of four 
or five miles from the station or sub- 
station presents no difficulties with 
this system, and the ability to take on 
business of this kind without exces- 
sive investment may be of considerable 
pecuniary benefit to the company. 

Careful attention must be given to 
the matter of balancing the load on 
long three-wire secondaries. The 
number of lights connected to each 
side may be nearly equal, but tests 
often show loads largely out of. bal- 
ance. Periodical tests with a portable 
cable-testing current transformer and 
ammeter are necessary in order to 
avoid trouble from this source. Dis- 
tribution at 220-440 volts has been 
resorted to by some suburban com- 
panies in order to be able to group 
a larger number of customers on one 
transformer and thus reduce distribu- 
tion losses. This scheme has been 
somewhat burdensome on the custom- 
ers, on account of the low efficiency 
of the 220-volt lamps, and has become 
even more unsatisfactory since the 
metallic-filament lamps have been 

placed upon the market, as they are 
not made at all in the higher voltages. 
No greater boon to the suburban com- 
panies could be devised than the de- 
velopment of a high-efficiency 220-volt 
lamp, as the large investment in small 
transformers and consequent high dis- 
tribution losses are serious obstacles 
to the extension of business in scat- 
tered districts. 


It is very desirable that circuits 
of different kinds be systematically ar- 
ranged on the cross-arms. This makes 
it easier to handle line trouble in 
emergency and also facilitates the 
keeping of records. Ideas differ as 
to the best arrangement, and no uni- 
form plan can be given that will apply 
to all conditions. In general the fol- 
lowing rules apply: 

Wires carrying the highest voltage 
should be on the upper cross-arms. 
Feeders and circuits to remote points 
should also be on the upper cross-arms, 
where they will not be disturbed. 
Primaries and street-lighting circuits 
for local use should be on the lower 
arms, where taps and transformer 
connections can be conveniently made. 
Secondary wires should be on the 
lowest arm by themselves. The ne- 
cessity for the last rule has been 
questioned, but if adhered to it un- 
doubtedly makes service and trouble- 
work safer, as it is sometimes difficult 
to distinguish between a three-phase 
primary and a three-wire secondary. 

The additional safety to life secured 
secondaries is now fully appreciated 
by grounding the neutral of three-wire 



July, 1908 

by everyone and is recommended by 
the Underwriters. Various methods 
of grounding the neutral have been 
used in the past. The most practical 
method is to drive a galvanized gas 
pipe into the earth at the foot of the 
pole. The ground wire is brought 
down the pole in wooden molding and 
connected to the ground pipe, which 
is also sometimes protected above the 
surface of the ground by a larger 
wooden molding. The problem of 
making a permanent joint between the 
ground wire and pipe has been a both- 
ersome one. Soldered joints and 
plugs have been tried with indifferent 
success. The ground cap devised by 
the engineers of the Commonwealth- 
Edison Company seems to have solved 
the problem. This consists of a malle- 
able galvanized-iron cap provided with 
an internal groove into which the No. 
6 ground wire is wedged between the 
inside of the cap and the ground pipe. 
The cap protects the end of the pipe 
while it is being driven and makes a 
rigid contact. Experience indicates 
that reliance should not be placed on 
a single ground, and a safe plan is to 
install at least two grounds on each 
secondary. It is possible that some- 
time in the future the practice will be- 
come standard of grounding the neu- 
tral wire of each house service to the 
water-piping of the house. 


Question. — We are going to buy a 
motor of 30 h.p., 3 ph., 60 cycles, 220 
volts, and the power company is forc- 
ing us to purchase transformers to 
step dozen the line voltage to our 
working pressure. Will three 10-h.p. 
(or say 7.5-&7C. ) transformers, or the 
equivalent of two lyh.p. transform- 
ers, be satisfactory to use? 

Answer. — No. Use one kilowatt in 
transformer capacity for every horse 
power in the motor. This is to give 
you ample current at starting when 
the power factor is low. In very 
small sizes, from about five horse 
power down, 1.5 kw. in transformers 
should be used for each horse power 
in the motors. 

Question. — We have never had sta- 
tion arresters on our lines, but some 
have been purchased for me to cut in. 
Should they be put on the line between 
the choke coil and the outside, or be- 
tween the coil and the generator? 

Answer.: — The proper connection is 
as follows: 

To Gt n t o a tq a. 

Choke Coiu 

The object of this arrangement is to 
throw back on the arresters, for 

ground discharge, any excessive volt- 
age which may reach the choke coils 
from outside sources. 

Question. — We have three exciters 
in one plant and two more arc to be 
installed. Will this mean additional 
Tirrell regulators? 

Answer. — The manufacturers of the 
Tirrell regulator make them to take 
care of as many as 12 exciters by 
working certain changes on the relays. 
You will have to consult them to find 
out what method to follow in your 
particular case. 

Question. — The armature of our 
generator got into trouble recently, 
and when I disconnected some of the 
leads from the commutator I found 
underneath a lot of zvires running 
around the armature. What could 
they have been for? 

Answer. — These were probably 
"equalizing connections." In any mul- 
tipolar generator there will be a num- 
ber of points of equal voltage for 
every pair of poles. It is practically 
impossible to get every pole exactly 
like every pole, so that one may be a 
better magnet than another. Accord- 
ingly, the armature might generate 
1 10.5 under one pair of poles and no 
under another. As the armature re- 
sistance is low, this one-half volt 
would be accompanied by heavy am- 
perage. This low voltage current 
would tend to circulate within the 
armature, causing excessive heating 
and sparking at the commutator. To 
overcome this, about every seventh 
bar of a commutator is connected to 
every other point of similar voltage 
by an equalizing connection. 

Question. — / have had to dig out 
the insulation between some of the 
bars on the commutator of our m-a- 
chinc. This causes continual trouble 
due to copper dust and compound ac- 
cumulating, and getting into the crev- 
ices. How can I overcome it? 

Answer. — Go into any paint store 
or chemical shop and buy some "wa- 
terglass," clean out the crevices and 
fill with the above. 

Question. — We arc soon to start up 
some high-tension transformers. The 
manufacturers haze sent instructions 
to dry them out thoroughly before 
starting up for the first time. As there 
is no standard of dryness, how am I 
to know when they are thoroughly 

Answer. — There is no way of telling 
when a transformer is thoroughly dry. 
All you can do is to take insulation 
resistance. As this would be lower 
with the transformer hot than cold, 
you should take it as nearly as possible 

to probable operating temperature. By 
referring to the standardization rules 
of the A. I. E. E., you can find out 
about what resistance you should look 

Question. — How can I detect the 
presence of zuater in my transformer 
oil; also acid? 

Answer. — Roast some blue stone 
until the water in it has been driven 
out and it has turned white. If the 
addition of some transformer oil will 
restore the blue color, it means that 
water is in the oil. Acid is most easily 
detected by using blue litmus paper, 
which turns red under the pressure of 

The "Western Electric Company 
Enters the Steam Turbine Field 

The Western Electric Company has 
recently closed a deal whereby it will 
manufacture and sell the Rateau 
turbine in connection with its genera- 

The turbines are very similar to the 
Rateau type as built in Europe, ex- 
cept that the construction is a little 
heavier and stronger to meet American 
conditions. They are of the impulse 
type and divided into a large number 
of stages. The wheels are turned out 
of a sojid steel plate, the cross-section 
of which gradually increases toward the 
center. The buckets are made of a spe- 
cial alloy of great mechanical strength 
and rust-resisting properties. The 
buckets are secured independently of 
each other bv means of special rivets 
of great mechanical strength, and are 
spaced at the periphery by a spacing 
ring, which serves to maintain an ac- 
curate spacing of the buckets, as well 
as to act as a baffle for improper cur- 
rents of steam. The turbines have 
fixed diaphragms which extend to the 
shaft between the wheels and form a 
cell in which each wheel operates. 

The governor is of the spring-bal- 
ance fly-ball type, operating in con- 
nection with a dash-pot, and is located 
in a cylindrical casing on the turbine 
bearing. The governor regulates-the 
speed by means of a double-beat type 
valve, which throttles the steam. 

The bearings are of the most simple 
construction, being practically the 
same as those of the ordinary ring- 
oiled dynamo, except that they have 
water jackets to maintain the tempera- 
ture at the desired value. The turbine 
being of the impulse type, there is 
practically no end thrust, and only a 
few thrust collars are necessary to lo- 
cate the shaft, these being placed at one 
end of the governor bearing. 

The Wisconsin Steel Company al- 
ready has in operation one of these 
turbines, operating on exhaust steam 
from the blooming engines and driving 
a direct-current generator. 

The National Electric Light Convention 

THE annual convention of the Na- 
tional Electric Light Associa- 
tion was held at the Auditorium 
Hotel, Chicago, on May 19th, 20th, 
2 1 st and 22d. The program was as 
follows : 

Tuesday, May 19th, Opening Ses- 
sion, 10 o'clock. — President's ad- 
dress ; announcements by the secre- 
tary ; report of committee on progress, 
Mr. T. Commerford Martin, New 
York; "Distribution in Suburban Dis- 
tricts," Mr. George H. Lukes, Chi- 
cago, 111.; "Tape," Mr. Paul Liipke, 
Trenton, N. J. ; report of committee 
on grounding secondaries, Mr. W. H. 
Blood, Jr., chairman, Boston, Mass. ; 
"Series Incandescent Lighting with 

Session, 10 o'Clock. — Report of com- 
mittee on gas engines, Mr. W. C. L. 
Eglin, chairman, Philadelphia, Pa. ; 
"Low-Pressure Steam Turbines," Mr. 
John W. Kirkland, Schenectady, N. 
Y. ; report of committee on meters, 
Mr. L. A. Ferguson, chairman, Chi- 
cago, 111. ; "Receiving Stations Oper- 
ated from High-Tension Lines," Mr. 
S. Q. Hayes, Pittsburg, Pa. ; report of 
committee on organization possibili- 
ties, Mr. Henry L. Doherty, New 
York City; special meeting, Banquet 
Hall, sixth floor. Parallel Session, 
10.15 o'clock. — "Uniform Accounting 
and Its Details," open to all account- 
ants and others interested ; Mr. H. M. 
Edwards, chairman committee on uni- 

guson, past president, Chicago, 111. 2. 
Preparation for a Campaign. (a) 
Field Work and Other Essentials ; (b) 
Analysis of Customers' Accounts; (c) 
Proportion of Lamp Equivalent Lost 
to Lamps Connected, — Showing Per- 
centage in Cities of Varied Popula- 
tion; (d) Policy of Handling Com- 
plaints; (e) Policy of Handling Col- 
lections. Editor, Mr. H. J. Gille, Min- 
neapolis, Minn. 3. The Contract 
Agent and the Representative, (a) 
The Contract Agent — His Possibili- 
ties; (b) The District Representative 
— His Possibilities; (c) The Special 
Representative: 1, The Sign Expert; 
2, The Power Expert ; 3, The Woman 
Representative; (d) Solicitors' Meet- 


Tungsten Lamps," Mr. P. D. Wag- 
oner, Schenectady, N. Y. Afternoon 
Session, 2.30 o'clock. — "Observations 
on the Precision of Different Types of 
Photometer," Prof. A. E. Kennelly 
and Mr. S. E. Whiting, Harvard Uni- 
versity ; "Power-Load Development 
for Central Stations of Moderate Size ; 
Some Unappreciated Possibilities," 
Mr. Charles Robbins and Mr. J. R. 
Bibbins, Pittsburg, Pa.; "The Small 
Station and its Economical Opera- 
tion," Mr. J. T. Whittlesey, Newark, 
N. J., and Mr. Paul Spencer, Phila- 
delphia, Pa. 

Wednesday, May 20th, Morning 

form accounting, in the chair. Eve- 
ning Session, 8 o'clock. — Executive 
session ; report of secretary and treas- 
urer and executive committee ; report 
of insurance expert, Mr. W. H. Blood, 
Jr., Boston, Mass. ; report of com- 
mittee on public policy, Mr. Arthur 
Williams, chairman, New York City ; 
"The Status and Commercial Possi- 
bilities of High-Efficiency Lamps" 
and Discussion, Mr. W. W. Freeman, 
Brooklyn, N. Y. 

Thursday, May 21st. — Commercial 
Day. — 1. Address — "Relationship Be- 
tween the Engineering and Commer- 
cial Departments," Mr. Louis A. Fer- 

ings — Their Objects. Editor, Mr. V. 
A. Henderson, Memphis, Tenn. 4. 
The Display Room, (a) Appoint- 
ments and Methods; (b) Value of 
Special Demonstrations; (c) Value 
of Electrical and Food-Show Exhib- 
its. Editor, Mr. L. G. Mathes, Du- 
buque, Iowa. 5. Advertising, (a) 
What is Being Done; (b) Why? (c) 
Results. Editor, Mr. Charles \. 
Parker, Detroit, Mich. 6. Publicity. 
(a) Methods to Create Proper Public 
Sentiment; (b) Dormant Publicity 
Opportunities of Lighting Companies. 
Editor, Mr. Percy Ingalls, Newark, 
N. J. 7. Creating Demand for Elec- 




July, 1908 

tricky, (a) The Creative Principle; 
(b) Notable Examples; (c) .Stereop- 
tioon Talk upon Outline and Sign 
Lighting, Showing Progress in Large 
and Small Cities. Editor, Mr. Frank 
B. Rae. Jr.. New York City. 8. Evo- 
lution of New-Business Building, (a) 
Examples of Central Stations that 
Have Continued Methods During De- 
pression ; (b) Strong Plea for Up- 
keep of Commercial Departments and 
Advertising; (c) Opportunities for 
Creating Business Along Existing 
Lines. Editor, Mr. George X. Tidd, 
Scranton. Pa. 9. The Electrical Con- 
tractor. Symposium; (a) What He is 
Doing to Assist in Creating Greater 
Demand for Electricity; (b) Specific 
examples. Editor. Mr. Joseph F. 
Becker. Jr.. Brooklyn. X. Y. 10. "Co- 
operative Commercialism,'" Mr. J. 

Among those present were the fol- 
lowing: J. B. Adams, Waterbury 
Company, Xew York City; Godfrey 
H. Atkin, Electric Storage Battery 
Company, Chicago ; Morgan Brooks, 
professor of electrical engineering. 
University of Illinois ; W. J. Barr, 
president Guaranty Electric Heater 
Company, Cleveland ; B. A. Behrend, 
chief electrical engineer Allis-Chal- 
mers Company, Milwaukee, Wis. ; Ja- 
cob Bonn, president Sangamo Electric 
Company, Springfield, 111. ; C. E. 
Brown, secretary Central Electric 
Company, Chicago ; D. J. Burns, gen- 
eral sales manager Ward Leonard 
Electric Companv. Bronxville, N. Y. ; 
T. H. Brady, T. H. Brady Manufac- 
turing Company. Xew Britain, Conn.; 
C. O. Baker, Baker & Company, Xew 
York City : Charles Blizard, third 


Robert Crouse, Cleveland, Ohio. II. 
"Illuminating Engineering as a Com- 
mercial Factor," illustrated, Mr. V. R. 
Lansingh, Xew York City. 12. Report 
of committee on solicitors' handbook 
prize award. Mr. John F. Gilchrist, 
chairman, Chicago, 111. 13. Report of 
committee on co-operative electrical 
development, Mr. W. W. Freeman, 
chairman, Brooklyn. X. Y. 

Friday, May 22<\. — "Illuminating 
Engineering." Mr. W. D'A. Ryan. 
West Lynn, Mass.: Report of com- 
mittee on protection from lightning 
and other static disturbances, Mr. R. 
S. Stewart, chairman, Detroit, Mich. ; 
"The Yalue of Care and Maintenance 
of Meters." Mr. H. D. King. Hoboken, 
X. J.: "Some Experiments in Com- 
bustion," Mr. S. J. Lenher, Xew 
York: "Specifications for Construc- 
tion on Joint Poles," Mr. Paul 
Spencer, Philadelphia, Pa. 

vice-president Electric Storage Bat- 
tery Company, Philadelphia : A. Ben- 
son, International Electric Meter 
Company, Chicago; C. E. Corrigan. 
vice-president National Metal Mold- 
ing Company, Pittsburg; William 
Coale, treasurer and manager Ster- 
ling Electric Manufacturing Com- 
pany. Warren, Ohio ; Frank J. Coak- 
ley, Samson Cordage Works, Boston ; 
Walter Cary, general manager West- 
inghouse Lamp Company, Xew York 
City; W. W. Cheney, jr., president 
International Electric Meter Com- 
pany, Chicago; C. A. Dubosch, man- 
ager Hugo Reisinger, Xew York 
City ; A. J. DeCamp, manager Phila- 
delphia Electric Company, Phila- 
delphia : S. E. Doane, chief engineer 
Xational Electric Lamp Association, 
Cleveland : F. L. Driver, president 
and treasurer Driver-Harris Wire 
Company, Xewark, N. J. ; J. H. Dale, 

president Dale Company, New York 
City; Avery P. Eckert, manager sales 
Duplex Metals Company, Xew York 
City; A. C. Garrison, president Co- 
lumbia Incandescent Lamp Company, 
St. Louis ; Rodman Gilder, publicity 
manager Crocker- Wheeler Company, 
Ampere, X. J.; F. H. Gale, in charge 
of advertising, General Electric Com- 
pany, Schenectady, X. Y. ; H. M. 
Hirschberg, president Excello Arc 
Lamp Company, Xew York City ; F. 
E. Hunting, sales manager and treas- 
urer Fort Wayne Electric Works, 
Fort Wayne, Ind. ; E. H. Haughton, 
manager Bryan-Marsh Company, Chi- 
cago; Alexander Henderson, Amer- 
ican Circular Loom Company, Bos- 
ton ; W. S. Heger, assistant to the 
president, Allis-Chalmers Company, 
Milwaukee ; W. H. Jacob, sales man- 
ager Triumph Electric Company, Cin- 
cinnati, Ohio; Basil G. Kodgbanoff, 
sales manager Benjamin Electric 
Manufacturing Company. Chicago ; A. 
X. Fox. advertising manager Ben- 
jamin Electric Manufacturing Com- 
pany. Chicago; B. C. Kenyon, presi- 
dent Diehl Manufacturing Company, 
Elizabethport, X. J. ; \*. R. Lansingh, 
chief engineer and general manager 
Holophane Company, Xew York 
City : R. C. Lamphier. secretary and 
manager Sangamo Electric Company, 
Springfield. 111. ; W. W. Low, presi- 
dent Electric Appliance Company. 
Chicago; W. A. Layman, vice-presi- 
dent and general manager Wagner 
Electric Manufacturing Company, St. 
Louis; R. K. Mickey, president Xov- 
elty Incandescent Lamp Company. 
Emporium. Pa. ; George A. McKin- 
lock, president Central Electric Com- 
pany, Chicago ; George T. Manson, 
general superintendent Okonite Com- 
pany, Limited. Xew York City ; Joseph 
E. Montague, general manager Buf- 
falo & Xiagara Falls Electric Light 
and Power Company, Xiagara Falls, 
X. Y. ; J. C. McQuiston, manager 
Westinghouse Companies' Publishing 
Department. Pittsburg: W. M. Mat- 
thews, treasurer W. X. Matthews & 
Brother. St. Louis; W. W. Merrill, 
secretary Chicago Fuse Wire and 
Manufacturing Company, Chicago ; 
X. L. Xorris. secretary and manager 
Banner Electric Company. Youngs- 
town, Ohio; Frank S. Price, secretary 
Pettingell-Andrews Company, Bos- 
ton ; H. M. Post, advertising manager 
Western Electric Company, Chicago; 
George F. Porter, sales manager At- 
lantic Insulated Wire and Cable Com- 
pany, Xew York City ; A. H. Patter- 
son, vice-president Phoenix Glass 
Company. Xew York City ; J. W. 
Perry, manager electrical department. 
H. W. Johns-Manville Company. 
Xew York City ; H. C. Rice, vice- 
president General Incandescent Lamp 
Company, Cleveland . 

; ffiK pi - 
^i2S£fw, mo. 


Volume XXXIX. Number 7. 
$1.00 a year; 15 cents a copy 

New York, August, 1 908 

The Electrical Age Co. 
New York and London 

Published monthly by The Electrical Age Co., 45 E. 42d Street, New York. 

J. H. SMITH. Pres. C. A. HOPE, Sec. and Treas. 


Telephone No. 6498 38th. Private branch exchange connecting all departments. 
Cable Address — Revolvable, New York. 


United States and Mexico, $1.00; 

Canada, $1.50. To Other Countries, $2.50 


Insertion of new advertisements or changes of copy cannot be guaranteed for the following 
asue if received later than the 15th of each month. 


Business and Editorial Departments, 33 Bedford Street 

Strand. W. C. 

London Telephone, Post Office, 3807 Central. Private branch exchange. London Cable 

Address Revolvable, London. 

All remittances to London Office should be crossed London and County Banking Company. 


Subscriptions may be sent to the following branch offices or agencies, where they will re- 
ceive the same careful attention as at the home office in New York : 

124, Queen St., Melbourne 33, Loop St., Cape Town 

359, George St., Sydney Unter den Linden, 5, Berlin 

61, Main St., Yokohama Nevsky Prospect, St. Petersburg 

23, Esplanada Rd., Bombay 31, bis rue de Fauburg Montmartre. Paris 

General Agents for United States and Canada : The American News Company 

The Development of the Regulat- 
ing Converter 

It is well known that a rotary con- 
verter maintains practically a definite 
and constant ratio between the electro- 
motive force at the collector rings and 
the electromotive force at the commu- 
tator. In order to vary the electro- 
motive force of the direct current it 
is therefore necessary to vary the 
electromotive force of the alternating 
current supplied to the machine. 
Various well-known methods are used 
for this purpose, such as a transform- 
er with variable ratio, made by chang- 
ing the number of turns in one 
of its windings. This requires 
a regulator with contact devices, 
which is in general objectionable, 
especially on large sizes, where the 
current is of considerable magnitude. 
Another form of regulator is the in- 
duction regulator, built somewhat sim- 
ilar to an induction motor ; the sec- 
ondary electromotive force delivered 
from the regulator depends upon rela- 
tive position of primary and sec- 
ondary. Consequently a change in the 
angular position of the secondary 
changes the electromotive force de- 
livered. This is used for varying the 
electromotive force supplied to the ro- 
tary converter. 

An induction regulator is a some- 
what cumbersome piece of apparatus, 
and ordinarily requires artificial cool- 

Another method which is in use for 
automatic compounding is the choke 
coil, which is sometimes placed in the 
alternating-current conductors. This 
method of control requires a varying 
power factor, which is often objection- 

The Synchronous Regulating 

A new method of control, which has 
lately been introduced in connection 

with a number of rotary converters in 
New York, employs an auxiliary 
alternator, mounted on the shaft of 
the rotary converter. This alternator 
has a number of stationary field poles, 
equal to the number of field poles of 
the rotary converter. The alternating- 
current conductors leading to the ro- 
tary converters are passed through the 
armature of the auxiliary alternator. 
The electromotive force induced in 
this auxiliary machine is under control 
of its field current. This electromotive 
force may, therefore, be made to in- 
crease that of the supply circuit, or it 
may be reversed by reversing the di- 
rection of the field current, in which 
case it will oppose the electromotive 
force of the supply circuit, and de- 
liver to the rotary converter a reduced 
electromotive force. The principle in- 
volved is therefore quite elementary 
and simple. The apparatus is likewise 
of a simple type, and is quite easily 

The action is substantially the same 
as that of an ordinary booster in con- 
nection with the direct-current ma- 

John A. Roebling 

The victories of the engineer over 
the forces and obstacles of nature are 
so recent in the world's history that 
people have not yet gotten the habit 
of regarding them as of equal im- 
portance to those of the soldier, the 
sailor and the statesmen. 

It is therefore genuine pleasure to 
note that the good people of Trenton 
have erected a monument to the mem- 
ory of John A. Roebling, of whom the 
late Charles Hewitt said : "His name 
is one known wherever a knowledge 
of science has gone, as perhaps the 
most successful engineer of the age. 
He was gifted with the ability to de- 
vise and execute with equal success, 
and hence deserved and received the 
just praise of the scientific world." 

Motor Control Systems 

The evolution of variable speed con- 
trol systems has been toward sim- 
plicity, as in other branches of the 
electrical field. It is not many years ago 
since the five-wire system for obtain- 
ing variable speed was considered very 
favorably. For economy of current 
consumption this method will prob- 
ably never be surpassed, as the use 
of different voltages between wires 
gave almost unlimited opportunity to 
run without the insertion of resistance. 
But economy of current consumption 
was very much offset by high first 
cost, high maintenance and frequent 
shut-down for repairs. 

This system soon worked down to 
the later and popular three-wire plan, 
wherein either two voltages, as 1 10 
and 220, became available, or 65 be- 
tween two wires, 155 between two 
others, and 220 between outsides. 
Either of the two latter is occasionally 
used now, but mostly the straight two- 
wire single-voltage system is adopted. 

In this system, if first cost is the 
main consideration, an armature re- 
sistance box is used with a maximum 
of 2:1 speed control. If more varia- 
tion is required and first cost still 
governs, then 2:1 by armature control 
and 50 per cent, more by field control 
is often used. But when running cost 
is the chief factor the entire range of 
speed control is obtained by the in- 
sertion of resistance into the field 
circuit only. By the use of the com- 
pensating, or inter-pole motor, a 6:1 
change may be had. but it is very rare 
for this range to be called for now. 
1 hough common enough in the first 
enthusiastic days of variable motor 
speed. Tt is seldom that over 3:1 is 
now asked by machinery makers, and 
this is obtained by simply making the 
motor with a field of slightly greater 
capacity than is required for constant 
speed service. 

Tn the case of alternating-current 
systems, the speed control is obtained 




August, 1908 

by inserting resistance into the rotor, 
or secondary circuit. This is electric- 
ally equivalent to the armature re- 
sistance method used with direct-cur- 
rent motors. It means, of course, that 
losses are large and current-carrying 
parts are big and heavy. 

At present there seems no economi- 
cal method in sight for alternating- 
current variable speed. The plan of 
trying to use frequency changers and 
other such additional machinery has 
often been suggested, but, as far as 
we know, never tried. 

Where the polyphase motor re- 
quires at least a three-conductor wir- 
ing plan the single-phase system re- 
quires only two. The current con- 
sumption is still heavy, however. 
There is quite a verbal fight on now 
between central-station managers as to 
the relative merits of a straight single- 
phase system for all purposes as 
against the two or three-phase system. 

The single-phase advocates point to 
the multiplicity of wiring in the poly- 
phase system and to the difficulty of 
keeping the phases balanced with load. 
The polyphase enthusiasts, on the 
other hand, point out the difficulty of 
building up a power load on a single- 
phase system, due to the inability of 
obtaining satisfactory single-phase 
motors for all operations and the high 
cost of these motors under the most 
favorable conditions. 

As the alternating-current systems 
have practically swept the country, 
there now being only 300 direct-cur- 
rent central stations left out of a total 
of 4800. it is reasonable to look for 
many changes and improvements in 
alternating-current variable speed sys- 
tems in the near future. 

HoKomo, Marion &». "Western 
Traction Company 

As an illustration of what may be 
accomplished in building up an exist- 
ing electric railway, lighting and 
power distributing' system, and at the 
same time strengthening the industrial 
position of an entire communitv, the 
management of the Kokomo, Marion 
& Western Traction Companv fur- 
nishes one of the most striking and 
instructive examples to be met with 
anywhere in the country. At the time 
ownership of the system was assumed 
bv the interests now in control, less 
than four vears aero, there was a small 
street railway and electric lighting 
plant, having no greater output than 
500 kw.. a few miles of tra^ka°"e 
wholly within the citv of Kokomo. and 
circuits containing several hundred arc 
and incandescent lamps. To-dav 2\- 
000 incandescent and 395 arc lamps in 

and about Kokomo are supplied with 
electric current and more than one- 
half of the Company's customers have 
various electric household devices; 
factories in the vicinity take upwards 
of 1000 h. p. daily in current for 
operating motors ; the street railway 
system has been extended to a track- 
age of ten miles, and a finely equipped 
interurban line of 28 miles in length 
extends from Kokomo, a city of about 
18,000 inhabitants, to Marion, hav- 
ing a population of 26,000. through 
several large towns situated in a rich, 
closely tilled agricultural country. 
Further extensions are projected to 
Terre Haute and Lafayette on the 
west, distant respectively 130 and 70 
miles from the eastern terminus of the 

Setting a MarKet Value on a 
"Water-Power Plant 

"The value of a water-power plant." 
says Mr. Charles T. Main, of Boston, 
"varies extremely with different con- 
ditions which govern the first cost, 
and with the character of the work 
done. The effect of the head, length 
of dam. length of canal, distance from 
canal to river, etc.. increase or de- 
crease the cost of construction. Very 
much better work is done in some 
places than in others, which increases 
the value and decreases the deprecia- 
tion, so that no general rule can be 
given to cover all cases. The plant 
must be considered not alone, but in 
connection with the privilege, each be- 
ing dependent upon the other, and 
each affecting the value of the other. 

"For the water-wheels themselves, 
the average life of the wheel is prob- 
ablv about twenty-five vears. while the 
casing might be allowed to outlive two 
wheels. Iron or steel penstocks, if 
taken care of. should last probably 100 
years, but wooden feeders under- 
ground will not last fifty years. 
Wooden flumes, gates and racks which 
are exposed to the weather will last 
about twenty vears. Some wooden 
dams have lasted a great many years, 
but thev are apt to get washed away 
in freshets. Stone dams, if properly 
designed and well built, will last for 
hundreds of vears. 

"The market value of the wheels 
would depend somewhat upon their 
efficiencv. independent of their physi- 
cal conditions : for it might pav to re- 
place them, if water is expensive, bv 
wheels of higher efficiencv. The ver- 
tical wheels with bevel gears will not 
TT-oduce as much net horse nower per 
cnbiV foot of water as the horizontal 
wheels : and with the horizontal wheels 
the extra expense and <"1an°er of 
breakage of gears is avoided." 

Power Station Lighting 

Power station lighting is one of the 
minor details in the design of a sta- 
tion, but it is nevertheless an im- 
portant one. MaxCollbohm, underthe 
title of "Notes on Power Station Light- 
ing,"cites some very definite objections 
to 25-cycle current lighting, but most 
engineers will not agree with him about 
the matter. He advocates the use of 
direct current from the auxiliary stor- 
age battery and an automatic switch 
for use with the lighting circuits. 
There are not many cases where it 
would pay to put in the separate bat- 
tery and motor-generator set for light- 
ing. Where such apparatus has been 
installed for operating the oil circuit- 
breakers, or as a reserve source for 
excitation, it is entirely feasible to do 
the lighting from this battery. The 
automatic switch for the lighting 
circuits is not a new suggestion, as at 
least one manufacturer has used such 
a switch with a spring-controlled au- 
tomatic release attachment. 

SovitHern W^ater Power 

The water powers of the Southern 
States are rapidly coming to rival 
those of New England, and their de- 
velopment is due in no small degree 
to the work of the United States Geo- 
logical Survey, which has for a num- 
ber of years been making systematic 
studies of the flow of the streams and 
the conditions which affect that flow. 

The work in Georgia has been car- 
ried on for more than a decade, dur- 
ing which period all the more impor- 
tant streams and many of the lesser 
ones have been measured many times, 
and records have been kept of daily* 
monthly and seasonal variations in 
their flow. Most of the data thus col- 
lected have been published from time 
to time, but so many of the reports 
are out of print or otherwise inacces- 
sible that Messrs. B. M. Hall and 
M. R. Hall, who have had charge 
of the work, have assembled all the 
data relating to the State in a report 
just issued by the Survey as Water- 
Supply Paper No. 197. In addition to 
descriptions of the streams, records of 
daily gauge heights and estimates of 
monthly flow, the report includes 
tabulated elevations of the surfaces 
of the streams at specified points, by 
means of which the fall of the streams 
can be estimated for use as power, 
and indicates available undeveloped 
sites. A simple formula for determin- 
ing the horse-power when fall and 
flow are known is also presented, and 
incidental descriptions of the topo- 
graphic and geologic features of the 
State are given. The paper is ready 
for distribution, and copies may be 
obtained without charge by applying 
to the Director of the United States 
Geological Survey, Washington, D. C. 

Protective Devices 


Commonwealth Edison Co., Chicago 

WITH the introduction of con- 
stant potential distribution for 
incandescent lighting, engi- 
neers were confronted with the neces- 
sity of providing some form of pro- 
tection for generators and circuits, 
which would save them from the ef- 
fects of an abnormal How of current 
when the circuit was accidentally 
crossed or short-circuited. This prob- 
lem was fairly well solved for the con- 
ditions met in the early stages of the 
art, but it has reasserted itself with 
each increase in voltage and with the 
development of power stations of very 
large capacity. New and different 
solutions have been found for each 
case and the problem is still a live one. 

In the direct-current plants which 
were first installed the circuits were 
operated at about no volts. The most 
natural means of securing protection 
to the apparatus was the use of a de- 
vice to automatically cut off the sup- 
ply of electricity when more current 
was drawn from the circuit than it 
could safely carry. The presence of 
an overload or short-circuit was thus 
indicated in a way which required 
prompt action in the correction of the 
difficulty. It was found that wires of 
lead, tin and similar soft metals hav- 
ing a low melting point had a rela- 
tively high electrical resistance. This 
combination of physical properties 
afforded means by which an auto- 
matic cut-out could be provided in the 
form of a fusible connection inserted 
at the proper point in the circuit. 
These no-volt circuits were therefore 
protected by the insertion of short 
pieces of soft wire, known as fuses, 
which were so arranged that the 
melted pieces could readily be re- 
placed after conditions on the circuit 
had been restored to normal. 

In another method which was more 
elaborate a solenoid connected in 
series with the circuit tripped a spring 
which opened the switch in case of 
overload. This was called a circuit 

The use of fuses for protection 
against overloads and short-circuits in 
lighting systems became universal be- 
cause of the simplicity and low cost 
of fuse renewals. 

The blocks used for the support of 
fuses have been of various types. The 
earliest forms were of wood, these 
being followed by slate and other 
stone and later by porcelain. In its 
primitive form the fuse consisted of a 
piece of lead wire secured under bind- 
ing screws at each end. The uncer- 
tainty of this form of contact re- 

sulted in fuses blowing when they 
should not, and tips of copper suit- 
ably slotted to fit the binding screws 
were added. The use of wood was 
abandoned on account of risk of fire 
from the arc caused by the melting of 
the fuse. The use of slate and porce- 
lain, while it eliminated the fire risk 
incident to the wood block, resulted 
in the chipping of the surface or the 
cracking of the block in case of the 
blowing of the fuse under short-cir- 
cuit with large amounts of power 
available. The insurance interests 
therefore forbade the use of porcelain 
for fuse blocks except where the fuse 
was enclosed, and required that where 
slate or marble was used a suitable 
barrier be placed between the ter- 
minals, the purpose of this barrier be- 
ing to hold the heat of the arc away 
from the surface of the block, and to 
reduce the tendency of the arc to burn 
the contacts and binding screws. This 
barrier raises the fuse about an inch 
from the surface of the block and is 
quite effective. 

The danger of fire from the flash 
which occurs at the melting of the 

it cannot be removed without the use 
of tools. This was found necessary 
on account of the likelihood of covers 
being left off. This form of fuse is 
one of the best and least expensive 
methods of protecting low voltage 
branch circuits carrying loads of 1500 
watts and under. 

The protection of lines carrying 
larger loads was not found satisfac- 
tory with the plug type of fuse as the 
explosive force was too great when 
direct short-circuits occurred. The 
copper-tipped fuse wire known as the 
link fuse serves this purpose econom- 
ically and is quite satisfactory for 
loads up to 50 kw. or more at low 
potentials. The link fuse, however, 
is unsafe unless enclosed in a fire- 
proof box and mounted on a slate base 
with a suitable barrier between the 

The danger arising from the use of 
open link fuses led to the develop- 
ment of a large variety of enclosed 
cartridge fuses. Most of these con- 
sist of a tube of fibrous material in 
which the fuse is mounted, and a 
filling around the tube of certain fire 

JTVTtvov 1 

Primary Cutouts 

Fig. 1 

fuse when mounted on an open block 
led Edison at an early date to devise 
a form of enclosed fuse which could 
be easily replaced without the use of 
tools and which could be refilled when 
blown at small expense. This fuse is 
the now very familiarly known Edison 
plug fuse. Originally glass was used 
as the insulating medium and the 
cover was made removable, but it is 
now made of porcelain instead of 
glass and the cover is attached so that 

resisting powders which absorb the 
vaporized metal when the fuse blows 
and smothers the arc. Connection is 
made at the ends by means of brass or 
copper terminals, the copper being 
used on the fuses designed for cur- 
rents of 50 amperes and upwards. 

The use of glass and other non-po- 
rous substances in place of the fibrous 
tube has not been successful, as 
the pressure generated by the vapor- 
ization of the fuse metal within the 




August, 1908 

tube must have means of escape and 
the rigidity of the glass and similar 
solids is such that the tube is very 
apt to be exploded when the fuse 
blows on short-circuit. The conceal- 
ment of the fuse wire within the tube 
makes necessary some device for in- 
dicating when the fuse has melted. 
This consists usually of a hole in the 
tube which permits a small portion of 
the arc to burn a paper covering, thus 
indicating at the surface that the fuse 
has melted. 

Fig. 2 

The co§t of installation and main- 
tenance of cartridge fuses is neces- 
sarily several times greater than that 
of the link fuses. This has greatly re- 
tarded their adoption for low poten- 
tial circuits, where' the Edison plug 
and copper-tipped link fuses are most 
common. On 250 to 600-voll power 
circuits, the use of cartridge fuses is 
quite general. 

Tt is an unfortunate condition too 
frequently found that where the de- 
signing enginer has provided a safe 
and effective equipment of protective 
apparatus of the cartridge type, the 
operating engineer or manager, who 
knows little of the proper care of the 
apparatus, permits its effectiveness to 
be destroyed by the use of temporary 
devices designed to keep the circuit 
going but to postpone the expense of 
renewing the fuse. In many such 
cases the designing engineer might 
better provide the cheaper installation 
of link fuses properly enclosed in 
boxes and have an installation which 
is not so likely to be abused. 

The best design of a method of pro- 
tection for a distributing system is 
necessarily a compromise between 
conflicting conditions. On the one 
hand the number and location of fuses 
should be such that the area affected 
by the occurrence of trouble should be 
as small as possible, while on the other 
hand the fuse is a weak point in the 
circuit, prone to operate when it 
should not, and therefore should not 
be multiplied unnecessarily. 

Tn distributing electricity by over- 
head lines over an area where the load 

is scattered so that mains are not 
interconnected as shown in Fig. 1, 
experience has demonstrated that the 
arrangement of fuses indicated pre- 
sents a reasonably satisfactory com- 
promise between the requirements of 
minimum area affected and minimum 
number of fuses. 

The less important branches are 
isolated from the trunk feeds when 
trouble occurs on them without dis- 
turbing the remaining branches. In 
case of trouble on the trunk feeds the 
circuit is opened automatically at the 
station. If it has burned itself clear, 
as often happens, the circuit is closed 
and service is resumed very promptly. 
If it has not cleared itself the trunk 
feeds must he successively opened at 
the feeder end by a trouble man, until 
the one in trouble is located. The 
remaining portions of the circuit are 
then put into service and the trouble 
is located on the affected main as soon 
as possible. It has been found unde- 
sirable to provide fuses at the points 
where the heavy mains leave the 
feeder, as they frequently blow when 
they should not due to temporary 
overloads, deterioration of contacts or 
to trouble on some small branch which 
is severe enough to cause both branch 
and main fuses to blow simultane- 

supply so arranged that the occurrence 
of trouble will cut out reasonably small 
districts. Fuses at each junction are 
unnecessary and involve more risk of 
trouble by blowing when they, should 
not, than value in protecting the line 
against extended interruption. The 
work of repair is relatively quick 
and it is therefore justifiable to risk 
larger areas than with low-tension 
underground lines. 

In direct-current underground low- 
tension networks, the work of section- 
alizing must be done with the highest 
degree of refinement, owing to the 
density of the load, and the corre- 
sponding seriousness of an interrup- 
tion when it occurs. 

Trouble on a distributing main or 
>ervice taken from it must be limited 
to the block in which it occurs, and if 
lines are carried on both sides of the 
street, it must be restricted to one side 
of the street. Trouble on an under- 
ground main is usually of such a 
nature that considerable time is re- 
quired to make repairs so that service 
may be resumed. For these reasons 
it is usual to place fuses in under- 
ground mains at all points where they 
are connected into the system, so that 
in case of trouble the section affected 
will cut itself out and avoid the spread 
of the trouble further. 

Fig. 3 

In overhead low-tension networks, 
using weatherproof wire, the danger 
of short-circuit is very slight, if the 
lines are properly maintained, and it 
is therefore usual to omit fuses ex- 
cept at a few important points of 

In the early Edison systems a junc- 
tion box was provided for under- 
ground work similar to that illus- 
trated in Fig. 2. The fuse clips were 
equipped with copper-tipped fuse- 
made of sheet fuse metal which pro- 

August, 1908 



duced a large amount of vapor when 
it blowed under short-circuit and were 
subject to depreciation which caused 
them to heat and blow unnecessarily 
at times. This difficulty was largely 
obviated later by the introduction of 
sheet copper fuses such as those 
shown in Fig. 3, which are now in 

It is therefore necessary to provide 
primary fuses at each transformer 
and a sufficient number of primary 
main fuses to properly sectionalize the 
load and limit the area affected. In 
general, the likelihood of trouble in 
transformers and primary mains 
should not be greater than in the 

In alternating-current systems the 
tendency of the arc to hold when once 
established by the blowing of the fuse 
is fortunately less than in direct- 
current. This is due to the reversal of 
current twice during each cycle which 
permits the terminals to partially cool, 
as the wave of current passes through 
the zero point. For this reason the 
use of fuses designed for 250 volts 
direct-current is permitted by the in- 
surance authorities for alternating- 
current circuits up to 440 volts. 

Transformers should be provided 
with primary fuses of such size that 
they will not blow unnecessarily, and 
it is not advisable to attempt to pro- 
tect transformers against ordinary 
overloads on this account. It is there- 
fore usual to provide primary fuses 
having about twice the normal rated 
capacity of the transformer. The fol-' 
lowing table represents common prac- 
tice on 2200-volt systems : 




I K 

















1 ■_' 
























general use. This greatly reduced the 
weight of metal required and there- 
fore the severity of the arc at the time 
of the blowing of a fuse. The section 
of the copper fuse at the point where 
fusion takes place is designed to carry 
its normal rated load without undue 
temperature rise and to fuse at about 
twice its normal rating. The types of 
junction box used in connection with 
modern cable systems are shown in 
Figs. 3 and 4. 

The feeders are fused at the point 
where they feed the network to pro- 
tect the network against trouble on 
the feeder. It is not usual in large 
systems to provide fuses on the feeder 
at the station bus, as the operator on 
duty can open the switch and discon- 
nect the feeder in case it is necessary. 
The likelihood of feeder fuses going 
out under emergency conditions when 
they should not makes it safer to de- 
pend upon the operator for protection 
against feeder trouble, which in any 
case is rare. 

In alternating-current underground 
low-tension networks supplied by 
low-tension feeders, the same general 
principles apply as in the case of 
direct-current systems. 

With networks supplied by primary 
mains and transformers, as in Fig. 5, 
the situation is more complicated. The 
system must be protected against fail- 
ure of transformers and primary 
mains, as well as of low-tension mains. 

secondary mains, if the subway type The type of fuse which has proved 

of transformer is used in drained most satisfactory for transformers up 

manholes. to 30 kw. is illustrated in Fig. 6. The 

The severity of the arc on circuits removable porcelain plug carries con- 

r> — 


JO 5 

: a — 





To S7 


~~" Secohoar/es 

Fig. 5 

operating at potentials of 500 volts 
and more necessitates the use of some 
form of enclosed fuse on such circuits. 
The blowing of a 50-ampere link fuse 
in a slate-lined iron box on a 500-volt 
direct-current circuit has been known 
to totally wreck the box and in some 
cases to hold on long enough to par- 
tially melt the iron case. 

tacts on which the fuse is mounted 
and the heat formed by the melting of 
the arc produces an expulsive action 
which blows out the arc. This form 
of fuse is very satisfactory with alu- 
minum as the fuse metal up to 15 am- 
peres at 2200 volts. Above 15 am- 
peres the enclosed fuse in some form 
is preferable. The common cartridge 



August, 1908 

type is satisfactory in capacities up to 
50 amperes at 2200 volts, except 
where it is subject to moisture. In 
such circumstances the arc-smother- 
ing filler absorbs moisture, and fails to 
perform its work properly when the 
fuse blows. It is not practicable to 
fully shield the cartridge fuse from 
moisture when it is installed out of 
doors or underground and the use of 
this type of fuse on general distribut- 

fuse metal in alternating-current sys- 

below 100 am- 

It has a higher 

terns in the capacities 
peres to some extent, 
fusing point than lead and tin, but 
lower than copper. Its conductivity 
being better than lead and tin, fuses 
made of this metal are smaller and 
produce less vapor when they blow. 
In the capacities below 10 amperes its 
use is likely to be accompanied by 
trouble on out-of-door work, on ac- 

Fig. r, 

ing systems is not wholly satisfactory. 
Various other types of fuses have 
been devised from time to time, many 
of which have never been generally 
used. Of those which have come into 
a limited use, there is one which con- 
sists of two lignum vita? blocks be- 
tween which the fuse is mounted. A 
hole is provided in one of the blocks 
to permit the ejection of the vaporized 
fuse. The use of lignum vita? pre- 
vents shattering and yet it is fireproof 
to a momentary flash. This fuse is 
serviceable up to 50 amperes at 2200 






5 Ainp 

ere Fus 


- \ 



Fig. 7 



Another form consists of a tube 
through which the fuse runs, which is 
open at one end so as to expel the 
gases violently when the fuse blows. 
The tube is of fibrous material and is 
not well adapted for out-of-door serv- 
ice. Various other forms of expul- 
sion-type fuses have been used, but 
not very generally. 

Aluminum has been employed as a 

count of gradual deterioration which 

causes fusing when no 



The action of enclosed fuses is in 
general somewhat more sensitive than 
open link fuses on account of the 
more restricted radiation of the en- 
closed fuse. 

The time required to cause a 5- 
ampere fuse to operate at different 
loads is illustrated in the curve of 
Fig. 7. This curve is typical of the 
action of fuses of all sizes, the ab- 
solute values varying with different 
types and capacities. 

The law governing the operation of 
fuses was worked out by Preece in 
[888. It may be stated thus: 

Current=(z\/((/ ) :! 
(/ being the diameter of the wire ex- 
pressed in inches. The value of the 
constant a is different for each metal. 
For copper it is 10,244. for aluminum 
7585, for lead 1 379, for tin 1X42, and 
for iron 3 148. 

For instance, with a No. 10 B. & S. 
copper wire having a diameter of 
0.102 inch, the fusing current is 

c= 10,244 X \/( - 102 ) 3 — 334aniperes 
The fusing current for some of the 
^mailer sizes of wire are as follows for 
copper and aluminum : 

Size Wire B. & S 











V(5)>. ' 



. 229 







Fusing Current Copper 











Fusing Current Alu- 











At voltages higher than 4000, the 
use of fuses is limited to the sizes 
below 30 amperes, and above 10,000 
volts they are used only for trans- 
formers of small capacity such as 
pressure transformers. 

At pressures up to 600 volts, car- 
tridge fuses are made for currents up 
to 000 amperes. The cost of renewals 
in the larger sizes is such, however, 
that it is usually preferable to employ 
circuit breakers rather than fuses. 

The operation of the fuse being de- 
pendent on the elevation of its tem- 
perature, it is apparent that the relia- 
bility of its performance on over- 
loads depends upon the ease with 
which heat may be radiated. This is 
not a factor in" case of short circuit, a^ 
the temperature rise is so rapid that 
radiation has no appreciable effect. 

Under normal load conditions the 
fuse may fail to carry its rated load 
because of insufficient opportunity for 
radiation or because of insufficient 
contact at its terminals, which adds to 
the heat instead of assisting in carry- 
ing it away. A fuse with a long 
length of wire between terminal clips 
will generally act at a lower current 
than one made of a short length, and 
a fuse mounted on lugs of liberal area 
will carry more than the same fuse 
connected to small lugs. 

Under circumstances where auto- 
matic cut-outs operate at frequent in- 
tervals and on circuits operating at 
high voltages or controlling loads ot 
100 kw. and upward, the circuit- 
breaker possesses certain advantages 
over the fuse as a means of protection. 

In general the use of circuit-break- 
er^ is expensive in first cost but in- 
expensive in operation, while the use 
of fuses requires a minimum first cost 
with a maintenance charge. 

In central station distributing sys- 
tems the load is usually not widely 
variable and protective devices are not 
called upon to act except in case of 
line trouble. The use of fuses is 
therefore generally preferable in a dis- 
tributing system except on those 
feeder and transmission lines which 
carry large loads at high voltages 
where the use of fuses is not feasible. 

On low potential circuits the circuit- 
breaker consists of a switch of a de- 
sign suitable to control the maximum 
load of the circuit, with which is com- 
bined a coil connected in series with 
the circuit and so arranged that it will 
lift a movable core and release a 
spring- actuated mechanism which 
opens the switch. This plunger is de- 
signed to operate whenever the cur- 
rent exceeds a predetermined value. 

Circuit-breakers are commonly de- 

August, 1908 



signed so that they may he adjusted to 

operate at any point between 80 and 
150 per cent, of their normal rated 
capacity. It has been found in prac- 
tice that a magnetizing force of about 
1000 ampere turns is ample for the 
operation of the tripping device. 

In alternating-current systems the 
design of the circuit-breaker is modi- 
fied somewhat, because of the fact that 
on such circuit-breakers a series trans- 

verse-current trip releases the spring 
mechanism which in turn opens the 

In electrical operation the power for 
both closing and opening the circuit is 
supplied through solenoids or motors. 
The larger sizes and higher voltage 
breakers, such as those shown in Fig. 
8, are usually controlled electrically, 
on account of the power required and 
because of the greater facility of oper- 

former may be installed at a conve- 
nient point in the main circuit and small 
wires carrying a few amperes may be 
led from the series transformer to the 
tripping coil of the circuit-breaker. 
( )n circuits operating at 2200 volts and 
over, the switch is commonly of a de- 
sign which breaks in oil. The use of 
the series transformer on such circuit 
serves the double purpose of providing" 
a small current for operating the trip- 
ping device and of insulating the 
mechanism from the high potential 

Circuit-breakers are designed for 
two general classes of service in dis- 
tribution systems, protection against 
overload or short circuit, and protec- 
tion from the reversal of the flow of 
energy. The first class is known as 
overload circuit-breakers, while the 
second is called reverse-current ap- 

The operating mechanism of the 
circuit-breaker is controlled by hand 
or electrically by solenoids. 

In hand-operated breakers, the 
power required to open the circuit is 
usually stored in springs during the 
act of closing. The overload or re- 

Fig. 8 

ation permissible with remote con- 
trolled switches. The latter feature is 
quite essential in large systems where 
the number of switches to be handled 
during an emergency demands a sys- 
tem of control by which an operator 
may work rapidly and without great 

Direct current is usually available 
in stations and substations from the 
exciter system, and is often safe- 
guarded by a battery, which assures 
control at all times. It is therefore 
quite universal to use direct current 
for the operation of electrically actu- 
ated breakers. With the exception of 
the three-phase four-wire system it is 
not usual to employ single-pole cir- 
cuit breakers in alternating-current 
systems, as all the poles must be 
opened in order to avoid electrostatic 
disturbances. In the four-wire system 
the use of the neutral wire prevents 
distortion of the phases when one pole 
is opened, and single-pole breakers are 
therefore used in such systems. 

With electrically controlled appa- 
ratus the protective device is really the 
relay which energizes the control cir- 
cuit. This consists in general of an 

alternating-current solenoid energized 
by the main current transformer of the 
circuit, the plunger of which closes the 
direct-current circuit and energizes 
the mechanism of the circuit-breaker 
as shown in Fig. 9. 

Other direct-current control circuits 
are provided for the use of the oper- 
ator in opening and closing the 
breaker under normal conditions. 

In order to prevent the operation of 
the circuit-breaker under momentary 
rushes of current, it is usual to design 
the relay for operation with an inverse 
time element. That is with the relay 
set to operate at 100 amperes after 10 
seconds, it will operate at about 300 
amperes in five seconds and almost in- 
stantaneously at 1000 amperes. This 
characteristic results in prompt action 
in opening the circuit under short- 
circuit, while reducing the liability of 
unnecessary interruption under minor 

This result is accomplished by 
damping devices of various kinds, 
such as dashpots and air bellows. The 
air bellows has proved the most satis- 
factory in view of its simplicity and 
reliability. Recent improvements made 
in the air-release valves have improved 
the action of the bellows type of relay 
on heavy overloads, so that their oper- 
ation on short circuits is more prompt 
and the damage reduced accordingly. 

The arrangements of relays on a 
feeder or transmission line must be 
such that the occurrence of a short- 
circuit between any two wires will 
open the breaker. On single-phase 
circuits one relay is sufficient to ac- 
complish this. On two-phase three- 
wire systems used for power only 


Fig. 9 

which are not grounded normally one 
relay in the middle wire may be made 
to protect all three. In distribution 
circuits carrying lighting and power 
it is preferable, however, to provide 
separate relays and circuit-breaker- 
for the two outer wires so that only 
one phase is interrupted in case of 
trouble which does not short-circuit 
both phases. This is also true of the 
four-wire two-phase system. In the 



August, 1908 

three-wire three-phase system without 
grounded neutral the occurrence of a 
short-circuit between any two wires 
interrupts service on all phases and re- 
lays are required in two wires so that 
at least one will open the circuit in 
case of trouble on either phase. The 
circuit-breaker is therefore a three- 
pole breaker. 

In the four-wire three-phase system, 
or in a three-wire three-phase system, 
having the neutral point of the gener- 

substations with duplicate transmis- 
sion lines, it is usual to have conditions 
similar to those shown in Fig. 10. It 
will be noted that each transmission 
line is protected at the main bus by 
overload relays, while each converter 
is protected by reverse current relays 
on the direct-current side and by over- 
load relays on the alternating-current 
side. The latter will, however, oper- 
ate on overload without regard to the 
direction oT flow of energy. 

£1 i 

t r 






Fig. io 

ator winding grounded, it is essential 
that relays be installed in each phase 
wire, since the occurrence of a ground 
on either phase conductor results in a 
short circuit. In the four-wire system 
only the relays on the phases affected 
come into action. In case of a ground 
on one phase the circuit-breaker on 
that phase opens without interrupting 
service on the other two phases. 

Where several lines or banks of 
transformers are operated in parallel 
or where rotary converters supply a 
network, it is necessary that the ap- 
paratus be protected against a reversal 
of the flow of energy in case of 

Various methods have been em- 
ployed to accomplish this result, most 
of which employ a pivoted arm actu- 
ated by windings so arranged that 
when the direction of the flow of en- 
ergy is reversed, the arm closes the 
operating circuit and opens the circuit- 

In a large system embodying several 

In case a fault develops at the point 
A on line I, current will be drawn 
from the main bus and from the 
direct-current network through the 
converters which are operating from 
line 2 at the time the fault develops. 
The overload relay on line I at the 
bus should at once cut off the supply 
from this source. The supply through 
the converters should also be cut off 
by the action of the reverse current 
relays on the direct-current side of the 
converters, and perhaps by the over- 

load relays on alternating-current.side. 

If the trouble is severe and the lines 
2 and 3 are comparatively short, it 
may also happen that the current 
drawn from the direct-current bus 
may have operated the overload relays 
on the lines 2 and 3 at the main bus 
or on the converters operating from 
them at the time of the accident. If 
so the entire equipment which was 
operating on the three lines is shut 
down until it can be determined by test 
which of the three cables is faulty. 
This requires some time and the re- 
sumption of service is usually not pos- 
sible in less than 10 to 15 minutes, and 
if convenient testing facilities are not 
provided much longer. The time limit 
of the various relays should be so ad- 
justed that the number of relays oper- 
ated in an emergencywill be a mini mum. 

In case the overload relay at the 
main bus on line 2 and but one reverse 
current relay on the converters carried 
by this line had opened, the remain- 
ing converter connected to line 2 
would continue to operate as an in- 
verted converter. This might cause it 
to race and go to pieces, causing great 
damage to machine and substation. It 
is therefore important that converters 
be further protected by speed limit de- 
vices which will operate the direct- 
current circuit-breaker whenever the 
speed exceeds a safe maximum value. 
Such devices usually operate on the 
principle of centrifugal force, and be- 
ing called upon to act rarely should be 
tested out at regular intervals. 

The protection of distributing sys- 
tems from the effects of lightning 
might properly be classed with other 
protective measures, but the discus- 
sion of the subject involves so many 
considerations that it is regarded as 
preferable to treat it separately. 

Printing Press Data. 

The Whitlock Company, manufac- 
turers of a well-known line of large 
printing presses, has recently added 
two new sizes to its list of machines. 
In this connection they have taken the